.NET: Python: Merge main into feature-durabletask branch (#3385)

* Python: Add factory pattern to concurrent orchestration builder (#2738) * Add factory pattern to concurrent orchestration builder * Update readme * Address AI comments * Fix unit tests * Fix import * Prevent multiple calls to set participants or factories * Add comments * Mitigate warnings * Fix mypy * Address comments * Address Copilot comments * Fix tests * Python: fix: GroupChat ManagerSelectionResponse JSON Schema for OpenAI Structured Outpu… (#2750) * fix: ManagerSelectionResponse JSON Schema for OpenAI Structured Output Strict Mode * refactor: install pre-commit then commit again * Capture file IDs from code interpreter in streaming responses (#2741) * .NET: [BREAKING] Prevent nulls in AIAgent property (#2719) * prevent nulls in AIAgent property * address feedback * code ql sm04598 (#2723) Co-authored-by: Mark Wallace <127216156+markwallace-microsoft@users.noreply.github.com> * .NET: Add Conversation State Sample (Step05) (#2697) * Initial plan * Add Agent_OpenAI_Step05_Conversation sample for conversation state management Co-authored-by: rogerbarreto <19890735+rogerbarreto@users.noreply.github.com> * Update Program.cs comment to accurately describe the sample Co-authored-by: rogerbarreto <19890735+rogerbarreto@users.noreply.github.com> * Update the code to use the ConversationClient more in line with the samples in OpenAI * Apply suggestions from code review Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Changing sample to use ChatClientAgent and conversationId in GetNewThread --------- Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com> Co-authored-by: rogerbarreto <19890735+rogerbarreto@users.noreply.github.com> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Bump AWSSDK.Extensions.Bedrock.MEAI from 4.0.4.7 to 4.0.4.11 (#2777) --- updated-dependencies: - dependency-name: AWSSDK.Extensions.Bedrock.MEAI dependency-version: 4.0.4.11 dependency-type: direct:production update-type: version-update:semver-patch ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> * Bump Azure.Identity from 1.17.0 to 1.17.1 (#2780) --- updated-dependencies: - dependency-name: Azure.Identity dependency-version: 1.17.1 dependency-type: direct:production update-type: version-update:semver-patch - dependency-name: Azure.Identity dependency-version: 1.17.1 dependency-type: direct:production update-type: version-update:semver-patch - dependency-name: Azure.Identity dependency-version: 1.17.1 dependency-type: direct:production update-type: version-update:semver-patch - dependency-name: Azure.Identity dependency-version: 1.17.1 dependency-type: direct:production update-type: version-update:semver-patch ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> * Bump Azure.AI.AgentServer.AgentFramework from 1.0.0-beta.4 to 1.0.0-beta.5 (#2778) --- updated-dependencies: - dependency-name: Azure.AI.AgentServer.AgentFramework dependency-version: 1.0.0-beta.5 dependency-type: direct:production update-type: version-update:semver-patch - dependency-name: Azure.AI.AgentServer.AgentFramework dependency-version: 1.0.0-beta.5 dependency-type: direct:production update-type: version-update:semver-patch - dependency-name: Azure.AI.AgentServer.AgentFramework dependency-version: 1.0.0-beta.5 dependency-type: direct:production update-type: version-update:semver-patch ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> * Python: added more complete parsing for mcp tool arguments (#2756) * added more complete parsing for mcp tool arguments * fixed mypy * added nonlocal model counter, and some fixes * fixes in naming logic * extracted json parsing function, added parametrized test and checked coverage * Python: Updated package versions (#2784) * Updated package versions * Small fix * Bump actions/checkout from 5 to 6 (#2404) Bumps [actions/checkout](https://github.com/actions/checkout) from 5 to 6. - [Release notes](https://github.com/actions/checkout/releases) - [Changelog](https://github.com/actions/checkout/blob/main/CHANGELOG.md) - [Commits](https://github.com/actions/checkout/compare/v5...v6) --- updated-dependencies: - dependency-name: actions/checkout dependency-version: '6' dependency-type: direct:production update-type: version-update:semver-major ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: Chris <66376200+crickman@users.noreply.github.com> * .NET: adds support for labels in edges, fixes rendering of labels in dot a… (#1507) * adds support for labels in edges, fixes rendering of labels in dot and mermaid, adds rendering of labels in edges * Update dotnet/src/Microsoft.Agents.AI.Workflows/Visualization/WorkflowVisualizer.cs Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * escaping edge labels, adding tests for labels containing strange characters that would break the diagram and enabling the previous signature so the API has backwards compatibility. * Unify label in EdgeData * Edge API adjustments, removed useless "sanitizer" * fixed test --------- Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> Co-authored-by: Jacob Alber <jaalber@microsoft.com> Co-authored-by: Chris <66376200+crickman@users.noreply.github.com> * Python: Added custom args and thread object to ai_function kwargs (#2769) * Added an example of using kwargs in ai_function * Added thread object to ai_function kwargs * Updated docs * Small fix * Added thread parameter filtering * Fix WorkflowAgent to include thread convo history. Enable checkpointing. (#2774) * Update OpenAIResponses.yaml to match AgentSchema (#2598) 1. Update `connection` child types -- `kind: ApiKey` to `kind: key` otherwise schema will fail: https://microsoft.github.io/AgentSchema/reference/apikeyconnection/ 2. Update `outputSchema`'s `PropertySchema` to be `kind` instead of `type` otherwise schema will fail: https://microsoft.github.io/AgentSchema/reference/propertyschema/ * Python: Remove warnings from workflow builder on not using factories (#2808) * Revert concurrent * Fix comments * Python: Filter framework kwargs from MCP tool invocations (#2870) * Filter framework kwargs from MCP tool invocations * Fixes * Python: Fix WorkflowAgent to emit yield_output as agent response (#2866) * Fix WorkflowAgent to emit yield_output as agent response * use raw_representation * Raw representation handling * Python: Use agent description in HandoffBuilder auto-generated tools (#2713) (#2714) ## Summary Enhanced `HandoffBuilder._apply_auto_tools` to use the target agent's description when creating handoff tools, providing more informative tool descriptions for LLMs. ## Changes - Modified `_apply_auto_tools` to extract `description` from `AgentExecutor._agent` when available - Updated iteration to use `.items()` for more efficient dict traversal - Handoff tools now use agent descriptions instead of generic placeholders ## Example Before: "Handoff to the refund_agent agent." After: "You handle refund requests. Ask for order details and process refunds." ## Testing - All handoff tests pass (20/20) - No breaking changes to existing API Fixes #2713 Co-authored-by: Evan Mattson <35585003+moonbox3@users.noreply.github.com> * Python: [BREAKING] Observability updates (#2782) * fixes Python: Add env_file_path parameter to setup_observability() similar to AzureOpenAIChatClient Fixes #2186 * WIP on updates using configure_azure_monitor * improved setup and clarity * fixed root .env.example * revert changes * updated files * updated sample * updated zero code * test fixes and fixed links * fix devui * removed planning docs * added enable method and updated readme and samples * clarified docstring * add return annotation * updated naming * update capatilized version * updated readme and some fixes * updated decorator name inline with the rest * feedback from comments addressed * Python: Fix middleware terminate flag to exit function calling loop immediately (#2868) * Fix middleware terminate flag to exit function calling loop immediately * Eliminating duck typing * Improve function exec result handling * Fix race condition * Fix mypy issues * Python: Fix context duplication in handoff workflows when restoring from checkpoint (#2867) * Fix context duplication in handoff workflows when restoring from checkpoint * Address Copilot PR review * .NET: Update to latest Azure.AI.*, OpenAI, and M.E.AI* (#2850) * Update to latest Azure.AI.*, OpenAI, and M.E.AI* Absorb breaking changes in Responses surface area * Update dotnet/samples/AgentWebChat/AgentWebChat.AgentHost/Utilities/ChatClientExtensions.cs * Update dotnet/samples/AgentWebChat/AgentWebChat.AgentHost/Utilities/ChatClientExtensions.cs * Update dotnet/samples/AgentWebChat/AgentWebChat.AgentHost/Utilities/ChatClientExtensions.cs * Update dotnet/samples/GettingStarted/AgentWithOpenAI/Agent_OpenAI_Step04_CreateFromOpenAIResponseClient/Program.cs Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Using patch to remove the model is necessary, updated the response client to actually use the the ForAgent --------- Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> Co-authored-by: Roger Barreto <19890735+rogerbarreto@users.noreply.github.com> * Bump actions/download-artifact from 6 to 7 (#2862) Bumps [actions/download-artifact](https://github.com/actions/download-artifact) from 6 to 7. - [Release notes](https://github.com/actions/download-artifact/releases) - [Commits](https://github.com/actions/download-artifact/compare/v6...v7) --- updated-dependencies: - dependency-name: actions/download-artifact dependency-version: '7' dependency-type: direct:production update-type: version-update:semver-major ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> * Bump actions/cache from 4 to 5 (#2861) Bumps [actions/cache](https://github.com/actions/cache) from 4 to 5. - [Release notes](https://github.com/actions/cache/releases) - [Changelog](https://github.com/actions/cache/blob/main/RELEASES.md) - [Commits](https://github.com/actions/cache/compare/v4...v5) --- updated-dependencies: - dependency-name: actions/cache dependency-version: '5' dependency-type: direct:production update-type: version-update:semver-major ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> * Bump actions/upload-artifact from 5 to 6 (#2860) Bumps [actions/upload-artifact](https://github.com/actions/upload-artifact) from 5 to 6. - [Release notes](https://github.com/actions/upload-artifact/releases) - [Commits](https://github.com/actions/upload-artifact/compare/v5...v6) --- updated-dependencies: - dependency-name: actions/upload-artifact dependency-version: '6' dependency-type: direct:production update-type: version-update:semver-major ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> * Python : Ollama Connector for Agent Framework (#1104) * Initial Commit for Olama Connector * Added Olama Sample * Add Sample & Fixed Open Telemetry * Fixed Spelling from Olama to Ollama * remove"opentelemetry-semantic-conventions-ai ~=0.4.13" since its handled in a different pr * Added Tool Calling * Finalizing test cases * Adjust samples to be more reliable * Update python/packages/ollama/agent_framework_ollama/_chat_client.py Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Update python/packages/ollama/pyproject.toml Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Update python/packages/ollama/tests/test_ollama_chat_client.py Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Update python/packages/ollama/agent_framework_ollama/_chat_client.py Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Improved Docstrings & Sample * Update python/packages/ollama/agent_framework_ollama/_chat_client.py Co-authored-by: Eduard van Valkenburg <eavanvalkenburg@users.noreply.github.com> * Integrate PR Feedback - Divided Streaming and Non-Streaming into independent Methods - Catch Ollama Validation Error - Add OTEL Provider Name - Checked Ollama Messages - Add Usage Statistics * Revert setting, so it can be none * Validate Message formatting between AF and Ollama * Catch Ollama Error and raise a ServiceResponse Error * Fix mypy error * remove .vscode comma * Add Reasoning support & adjust to new structure * Add Ollama Multimodality and Reasoning * Add test cases for reasoning * Add Tests for Error Handling in Ollama Client * Update python/samples/getting_started/multimodal_input/ollama_chat_multimodal.py Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Integrated Copilot Feedback * Implement first PR Feedback * Adjust Readme files for examples * Adjust argument passing via additional chat options * Implemented PR Feedback * Removing Ollama Package from Core and moving samples * Fix Link & Adding Samples to Main Sample Readme * Fixing Links in Readme * Moved Multimodal and Chat Example * Fixed Link in ChatClient to Ollama * Fix AgentFramework Links in Ollama Project * Fix observability breaking change --------- Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> Co-authored-by: Eduard van Valkenburg <eavanvalkenburg@users.noreply.github.com> * Skip failing IT (#2904) * .NET: Cosmos DB UT Fast Skip (For Non-Configured Local envs) (#2906) * Cosmos DB UT Fast Skip (Non-Configured Local envs) + Long running UT skip in pipeline when no CosmosDB changes happened * Force a CosmosDB source code change to trigger the pipeline * Address possible string boolean mismatch * Add debug * Enabling emulator always when running IT * .NET: Add TTLs to durable agent sessions (#2679) * .NET: Add TTLs to durable agent sessions * Remove unnecessary async * PR feedback: clarify UTC * PR feedback: limit minimum signal delay to <= 5 minutes * PR feedback: Fix TTL disablement * Linter: use auto-property * Fix build break from OpenAI SDK change * Updated CHANGELOG.md * PR feedback * Reduce default TTL to 14 days to work around DTS bug * Python: Update Mem0Provider to use v2 search API `filters` parameter (#2766) * short fix to move id parameters to filters object * added tests * small fix * mem0 dependency update * Updated package versions (#2913) * .NET: Switch to new "Run" method name. (#2843) * Switch to new "RunAgent" method name. * Try to disable false positive naming warning. * Add comment about disabled warnings. * Rename `RunAgent` to just `Run`. * Update CHANGELOG. * Python: Switch to new "run" method name. (#2890) * Switch to `run` method. * Add support for deprecated `run_agent`. * Fix entity method name. * Fix method name and improve tests. * Update comment. * Update Python CHANGELOG. * [BREAKING] Python: Add factory pattern to handoff orchestration builder (#2844) * WIP: Factory pattern to handoff * Add factory pattern to concurrent orchestration builder; Next: tests and sample verification * Add tests and improve comments * Fix mypy * Simplify handoff_simple.py * Simplify handoff_autonoumous.py and bug fix * Update readme * Address Copilot comments * Python: Flow custom kwargs to agents via Workflow SharedState (#2894) * Flow custom kwargs to agents via SharedState * Address Copilot feedback * Improve sample typing * Fix test * Fix Pydantic error when using Literal type for tool params (#2893) * Updated Ollama package version (#2920) * Python: Azure AI Agent with Bing Grounding Citations Sample (#2892) * bing grounding sample with citations * small fix * fix * .NET: Make DelegatingAIAgent abstract (#2797) * Initial plan * Make DelegatingAIAgent abstract Co-authored-by: SergeyMenshykh <68852919+SergeyMenshykh@users.noreply.github.com> --------- Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com> Co-authored-by: SergeyMenshykh <68852919+SergeyMenshykh@users.noreply.github.com> * Added additional arguments for Azure AI agent (#2922) * Python: Correction of MCP image type conversion in _mcp.py (#2901) * Correction of MCP image type conversion in _mcp.py * Added a new overload to the init function of the DataContent() type of the Agent Framework, edited the test case to correctly test the usage of the data and uri fields while using DataContent() * Fixed tests related to the changes of the DataContent type, added testing for both string and byte representations * Pass kwargs into subworkflows (#2923) * Python: Move ollama samples to samples getting started dir (#2921) * Move ollama samples to samples getting started dir * Address feedback * Python: fix: correct BadRequestError when using Pydantic model in response_fo… (#1843) * fix: correct BadRequestError when using Pydantic model in response_format * Fix lint --------- Co-authored-by: Evan Mattson <evan.mattson@microsoft.com> * .NET: [Breaking] Delete display name property (#2758) * delete the AIAgent.DisplayName property * use agent name as a first value for activity display name * Update dotnet/src/Microsoft.Agents.AI.Workflows/Specialized/HandoffAgentExecutor.cs Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Python: cleanup and refactoring of chat clients (#2937) * refactoring and unifying naming schemes of internal methods of chat clients * set tool_choice to auto * fix for mypy * added note on naming and fix #2951 * fix responses * fixes in azure ai agents client * Python: Workflow add option to visualize internal executors (#2917) * Workflow add option to visualize internal executors * Address Copilot comments * Python: Fixes Run ID and Thread ID casing to align with AG-UI Typescript SDK (#2948) * added camelCase input to run id and thread id aligning with @ag-ui/core * fixed per copilot suggestions * Python: Add workflow cancellation sample (#2732) * Add workflow cancellation sample Add sample demonstrating how to cancel a running workflow using asyncio tasks. Shows both cancellation mid-execution and normal completion paths. Useful for implementing timeouts, graceful shutdown, or A2A executors. * update docstring * .NET: Update Anthropic package to version 12.0.0 (#2914) * Initial plan * Update Anthropic package to version 12.0.0 Co-authored-by: stephentoub <2642209+stephentoub@users.noreply.github.com> --------- Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com> Co-authored-by: stephentoub <2642209+stephentoub@users.noreply.github.com> * Python: Add Azure Managed Redis Support with Credential Provider (#2887) * azure redis support * small fixes * azure managed redis sample * fixes * Bump CommunityToolkit.Aspire.OllamaSharp from 13.0.0-beta.440 to 13.0.0 (#2856) --- updated-dependencies: - dependency-name: CommunityToolkit.Aspire.OllamaSharp dependency-version: 13.0.0 dependency-type: direct:production update-type: version-update:semver-patch ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> * Bump AWSSDK.Extensions.Bedrock.MEAI from 4.0.4.11 to 4.0.5 (#2853) --- updated-dependencies: - dependency-name: AWSSDK.Extensions.Bedrock.MEAI dependency-version: 4.0.5 dependency-type: direct:production update-type: version-update:semver-patch ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: Mark Wallace <127216156+markwallace-microsoft@users.noreply.github.com> * Bump Azure.AI.AgentServer.AgentFramework from 1.0.0-beta.4 to 1.0.0-beta.5 (#2854) --- updated-dependencies: - dependency-name: Azure.AI.AgentServer.AgentFramework dependency-version: 1.0.0-beta.5 dependency-type: direct:production update-type: version-update:semver-patch - dependency-name: Azure.AI.AgentServer.AgentFramework dependency-version: 1.0.0-beta.5 dependency-type: direct:production update-type: version-update:semver-patch ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: Chris <66376200+crickman@users.noreply.github.com> * Python: Fix WorkflowAgent event handling and kwargs forwarding (#2946) * Fix kwargs propagation through workflow.as_agent() * Fix WorkflowAgent to respect AgentExecutor output_response setting * .NET: Use GrpcEntityRunner instead of TaskEntityDispatcher (#2759) * Use GrpcEntityRunner instead of TaskEntityDispatcher * Pin to Durable worker 1.11.0 * Set the invocation result * Update all Durable packages * Update changelog, rename dispatcher to encondedEntityRequest * Python: Bump Py version to 1.0.0b251218 for a release. Update CHANGELOG (#2968) * Bump Py version to 1.0.0b251218 for a release. Update CHANGELOG * update lock * Fix formatting * Fix ChatKit typing * Python: Introducing Foundry Local Chat Clients (#2915) * redo foundry local chat client * fix mypy and spelling * better docstring, updated sample * fixed tests and added tests * small sample update * Updated package versions (#2978) * Python: Added GitHub MCP sample with PAT (#2967) * added github mcp sample with PAT * addressed copilot fixes * env fix * Python: Preserve reasoning blocks with OpenRouter (#2950) * Preserve reasoning blocks with OpenRouter * Put encrypted reasoning in TextReasoningContent * Remove unneccessary change * Fix docs * Support streaming * Fix handling None in TextReasoningContent.text * Python: Added response.created and response.in_progress event process to OpenAIBaseResponseClient (#2975) * added response.created and response.in_progress to include response.id * better doc string * added tests for the new streaming event types * Python: Introducing support for Bedrock-hosted models (Anthropic, Cohere, etc.) (#2610) * Pushing the bedrock related changes to the new branch after addressing the review comments * 2524 Addressed the second round review comments * 2524 Addressed few more minor comments on the PR * resolving the merge conflict * 2524 resolved the uv.lock conflicts * 2524 addressed more comments * 2524 removed the print statement to fix the checks failure * 2524 resolved the CI failure issues * 2524 fixing the CI breaks * 2524 Addressed the review comment * 2524 resolved conflict --------- Co-authored-by: Sunil Dutta <sunil.dutta@penske.com> Co-authored-by: budgetboardingai <apurva.sharma31@gmail.com> * .NET: [Durable Agents] Reliable streaming sample (#2942) * .NET: [Durable Agents] Reliable streaming sample * Add automated validation for new sample * Address Copilot PR feedback * Fix typo in README.md about agent definitions (#2634) * Fix typo in README.md about agent definitions * Update agent-samples/README.md Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> --------- Co-authored-by: Evan Mattson <35585003+moonbox3@users.noreply.github.com> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Python: latency improvements (#3014) * latency improvements * fixed mypy, added coding standards and instructions * slight logic improvement * Python: Updated package versions (#3024) * Updated package versions * Updated changelog * Python: add powerfx safe mode (#3028) * add powerfx safe mode * improved docstring and aligned env_file loading * ensured test uses reset * .NET: [Breaking] Introduce RunCoreAsync/RunCoreStreamingAsync delegation pattern in AIAgent (#2749) * Initial plan * Refactor AIAgent: Make RunAsync and RunStreamingAsync non-abstract, add RunCoreAsync and RunCoreStreamingAsync Co-authored-by: SergeyMenshykh <68852919+SergeyMenshykh@users.noreply.github.com> * Fix infinite recursion in test implementations Co-authored-by: SergeyMenshykh <68852919+SergeyMenshykh@users.noreply.github.com> * Make RunAsync and RunStreamingAsync non-virtual as requested Co-authored-by: SergeyMenshykh <68852919+SergeyMenshykh@users.noreply.github.com> * Fix DelegatingAIAgent subclasses to use RunCoreAsync/RunCoreStreamingAsync Co-authored-by: SergeyMenshykh <68852919+SergeyMenshykh@users.noreply.github.com> * Fix XML documentation references in AnonymousDelegatingAIAgent Co-authored-by: SergeyMenshykh <68852919+SergeyMenshykh@users.noreply.github.com> * Restore <see cref> tags with proper qualified signatures in AnonymousDelegatingAIAgent Co-authored-by: SergeyMenshykh <68852919+SergeyMenshykh@users.noreply.github.com> * Rollback unnecessary XML documentation changes in AnonymousDelegatingAIAgent Co-authored-by: SergeyMenshykh <68852919+SergeyMenshykh@users.noreply.github.com> * Remove pragma and update crefs to RunCoreAsync/RunCoreStreamingAsync Co-authored-by: SergeyMenshykh <68852919+SergeyMenshykh@users.noreply.github.com> * Fix EntityAgentWrapper to call base.RunCoreAsync/RunCoreStreamingAsync Co-authored-by: SergeyMenshykh <68852919+SergeyMenshykh@users.noreply.github.com> * fix compilation issues * fix compilatio issue * fix tests * fix unit tests * fix unit test --------- Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com> Co-authored-by: SergeyMenshykh <68852919+SergeyMenshykh@users.noreply.github.com> Co-authored-by: SergeyMenshykh <sergemenshikh@gmail.com> Co-authored-by: Chris <66376200+crickman@users.noreply.github.com> * add issue template and additional labeling (#3006) * fix and extra int test (#3037) * .NET: [BREAKING] Refactor ChatMessageStore methods to be similar to AIContextProvider and add filtering support (#2604) * Refactor ChatMessageStore methods to be similar to AIContextProvider * Fix file encoding * Ensure that AIContextProvider messages area also persisted. * Update formatting and seal context classes * Improve formatting * Remove optional messages from constructor and add unit test * Add ChatMessageStore filtering via a decorator * Update sample and cosmos message store to store AIContextProvider messages in right order. Fix unit tests. * Update Workflowmessage store to use aicontext provider messages. * Apply suggestions from code review Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Apply suggestions from code review Co-authored-by: SergeyMenshykh <68852919+SergeyMenshykh@users.noreply.github.com> * Improve xml docs messaging * Address code review comments. * Also notify message store on failure --------- Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> Co-authored-by: SergeyMenshykh <68852919+SergeyMenshykh@users.noreply.github.com> * [BREAKING] Remove unused AgentThreadMetadata (#3067) * Remove unused AgentThreadMetadata * Update DurableTask Changelog * Python: Fix AzureAIClient failure when conversation history contains assistant messages (#3076) * Fix AzureAIClient failure when conversation history contains assistant messages * Address PR review feedback: improve docstring and test assertions * Remove redundant cast * Fix: Update OTLP exporter protocol conditions (#3070) * Python: Fix ExecutorInvokedEvent and ExecutorCompletedEvent observability data (#3090) * Fix ExecutorInvokedEvent.data mutation bug * Fix bug related to not yielding output type * .NET: Seal ChatClientAgentThread (#2842) * Initial plan * Seal ChatClientAgentThread class Co-authored-by: SergeyMenshykh <68852919+SergeyMenshykh@users.noreply.github.com> --------- Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com> Co-authored-by: SergeyMenshykh <68852919+SergeyMenshykh@users.noreply.github.com> * Fix broken strands urls. (#3102) * Fix broken strands urls. * Fix typos * .NET: Fix message ordering inconsistency when using AIContextProvider (#2659) * Initial plan * Fix message ordering inconsistency when using AIContextProvider Co-authored-by: westey-m <164392973+westey-m@users.noreply.github.com> * Revert to original message ordering: Input, AIContextProvider, Response Co-authored-by: westey-m <164392973+westey-m@users.noreply.github.com> * Reorder messages to ChatClient to match MessageStore order: Existing, Input, AIContextProvider Co-authored-by: westey-m <164392973+westey-m@users.noreply.github.com> * Remove redundant test methods as existing tests already verify the behavior Co-authored-by: westey-m <164392973+westey-m@users.noreply.github.com> --------- Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com> Co-authored-by: westey-m <164392973+westey-m@users.noreply.github.com> Co-authored-by: Mark Wallace <127216156+markwallace-microsoft@users.noreply.github.com> Co-authored-by: Chris <66376200+crickman@users.noreply.github.com> * fix: tool_choice parameter not being honored when passed to agent.run() (#3095) * sharepoint sample fix (#3108) * Bump versions to 1.0.0b260106 for a release. Update CHANGELOG.md (#3109) * Bump Bedrock version to latest (#3110) * Python: Fix MCP tool result serialization for list[TextContent] (#2523) * Fix MCP tool result serialization for list[TextContent] When MCP tools return results containing list[TextContent], they were incorrectly serialized to object repr strings like: '[<agent_framework._types.TextContent object at 0x...>]' This fix properly extracts text content from list items by: 1. Checking if items have a 'text' attribute (TextContent) 2. Using model_dump() for items that support it 3. Falling back to str() for other types 4. Joining single items as plain text, multiple items as JSON array Fixes #2509 * Address PR review feedback for MCP tool result serialization - Extract serialize_content_result() to shared _utils.py - Fix logic: use texts[0] instead of join for single item - Add type annotation: texts: list[str] = [] - Return empty string for empty list instead of '[]' - Move import json to file top level - Add comprehensive unit tests for serialization * Address PR review feedback: fix type checking and double serialization - Add isinstance(item.text, str) check to ensure text attribute is a string - Fix double-serialization issue by keeping model_dump results as dicts until final json.dumps (removes escaped JSON strings in arrays) - Improve docstring with detailed return value documentation - Add test for non-string text attribute handling - Add tests for list type tool results in _events.py path * Simplify PR: minimal changes to fix MCP tool result serialization Addresses reviewer feedback about excessive refactoring: - Reset _events.py to original structure - Only add import and use serialize_content_result in one location - All review comments addressed in serialize_content_result(): - Added isinstance(item.text, str) check - Use model_dump(mode="json") to avoid double-serialization - Improved docstring with explicit return value documentation - Empty list returns "" instead of "[]" * Refactor: Move MCP TextContent serialization to core prepare_function_call_results Per reviewer feedback, moved the TextContent serialization logic from ag-ui's serialize_content_result to the core package's prepare_function_call_results function. Changes: - Added handling for objects with 'text' attribute (like MCP TextContent) in _prepare_function_call_results_as_dumpable - Removed serialize_content_result from ag-ui/_utils.py - Updated _events.py and _message_adapters.py to use prepare_function_call_results from core package - Updated tests to match the core function's behavior * Fix failing tests for prepare_function_call_results behavior - test_tool_result_with_none: Update expected value to 'null' (JSON serialization of None) - test_tool_result_with_model_dump_objects: Use Pydantic BaseModel instead of plain class * Fix B903 linter error: Convert MockTextContent to dataclass The ruff linter was reporting B903 (class could be dataclass or namedtuple) for the MockTextContent test helper classes. This commit converts them to dataclasses to satisfy the linter check. * Python: Improve DevUI, add Context Inspector view as new tab under traces (#2742) * Improve DevUI, add Context Inspector view as new tab under traces * fix mypy errors * fix: Handle stale MCP connections in DevUI executor MCP tools can become stale when HTTP streaming responses end - the underlying stdio streams close but `is_connected` remains True. This causes subsequent requests to fail with `ClosedResourceError`. Add `_ensure_mcp_connections()` to detect and reconnect stale MCP tools before agent execution. This is a workaround for an upstream Agent Framework issue where connection state isn't properly tracked. Fixes MCP tools failing on second HTTP request in DevUI. fixes #1476 #1515 #2865 * fix #1572 report import dependency errors more clearly * Ensure there is streaming toggle where users can select streaming vs non streaming mode in devui . Fixes .NET: [Python] DevUI tool call rendering in non-streaming mode? * remove unused dead code * improve ux - workflows with agents show a chat component in execution timelien, also ensure magentic final output shows correctly * update ui build * update devui to use instrumentation instead of tracing, other instrumentation and type/instance check fixes * .NET: Seal factory contexts and add non JSO deserialize overloads (#3066) * Seal factory contexts and add non JSO deserialize overloads * Apply suggestions from code review Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Enable blank issues in issue template configuration Need to re-enable creating blank issues * updated templates (#3106) * updated templates * enabled blank and fixed triage * made language optional and moved to the bottom for features * Python: Streaming sample for azurefunctions (#3057) * Streaming sample for azurefunctions * Fixed links and sample name * Addressed feedback * Addressed feedback * Fixed integration tests * Updated test * Python: fix(azure-ai): Fix response_format handling for structured outputs (#3114) * fix(azure-ai): read response_format from chat_options instead of run_options * refactor: use explicit None checks for response_format * Fix mypy error * Mypy fix * Python: Bump python version to 1.0.0b260107 for a release (#3128) * Bump python version to 1.0.0b260107 for a release * Update changelog * Make A2AAgent public, so that it's concrete implementation methods can be used. (#3119) * .NET: Map additional props <-> A2A metadata (#3137) * map additional props from agent run options to a2a request metadata * small touches * add unit tests for new extension methods * Sort using * add unit test * add additiona unit tests * special case json element to avoid unnecessary serialization * Python: Fix Anthropic streaming response bugs (#3141) * test commit identity * fix(anthropic): fix raw_representation and finish_reason in streaming * lint fix * Bump AWSSDK.Extensions.Bedrock.MEAI from 4.0.5 to 4.0.5.1 (#2994) --- updated-dependencies: - dependency-name: AWSSDK.Extensions.Bedrock.MEAI dependency-version: 4.0.5.1 dependency-type: direct:production update-type: version-update:semver-patch ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: Chris <66376200+crickman@users.noreply.github.com> * Bump Anthropic from 12.0.0 to 12.0.1 (#2993) --- updated-dependencies: - dependency-name: Anthropic dependency-version: 12.0.1 dependency-type: direct:production update-type: version-update:semver-patch ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: Chris <66376200+crickman@users.noreply.github.com> * .NET: [Breaking] Prevent loss of input messages & streamed updates when resuming streaming (#2748) * save input messages and stream updates to the continuation token to be able to use them in the last successful stream resumption call. * Update dotnet/src/Microsoft.Agents.AI/ChatClient/ChatClientAgentContinuationToken.cs Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Update dotnet/src/Microsoft.Agents.AI/ChatClient/ChatClientAgentContinuationToken.cs Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Update dotnet/tests/Microsoft.Agents.AI.UnitTests/ChatClient/ChatClientAgent_BackgroundResponsesTests.cs Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Update dotnet/src/Microsoft.Agents.AI/ChatClient/ChatClientAgentContinuationToken.cs Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Update dotnet/src/Microsoft.Agents.AI/ChatClient/ChatClientAgentContinuationToken.cs Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * fix typo * init continuation token from chat response * remove unnecessary types for source generation * remove check for continuation token passed at initial run * remove check for continuation token pass at initial run * centralize continuation token parsing * update xml comments * use readonly collection instead of enumerable --------- Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * .NET: fix: Expose WorkflowErrorEvent as ErrorContent (#2762) * fix: Expose WorkflowErrorEvent as ErrorContent When hosted using .AsAgent(), Workflows were not exposing inner errors coming as Exceptions (through the WorkflowErrorEvent) The fix is to convert their message to an ErrorContent on the way out, rather than rely on the default "empty update" to collect the raw event. * feat: Add a way to show/suppress exception information * Bump Microsoft.Agents.AI.Workflows from 1.0.0-preview.251125.1 to 1.0.0-preview.251219.1 (#2997) --- updated-dependencies: - dependency-name: Microsoft.Agents.AI.Workflows dependency-version: 1.0.0-preview.251219.1 dependency-type: direct:production update-type: version-update:semver-patch ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: Chris <66376200+crickman@users.noreply.github.com> Co-authored-by: SergeyMenshykh <68852919+SergeyMenshykh@users.noreply.github.com> * .NET: Add Run overloads to expose ChatClientAgentRunOptions in IntelliSense (#3115) * Initial plan * Add ChatClientAgentExtensions for improved discoverability of ChatClientAgentRunOptions Co-authored-by: westey-m <164392973+westey-m@users.noreply.github.com> * Address code review feedback - use collection expression syntax Co-authored-by: westey-m <164392973+westey-m@users.noreply.github.com> * Apply suggestion from @westey-m * Fix issues with Copilot implementation * Add additional tests for structured output overloads. --------- Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com> Co-authored-by: westey-m <164392973+westey-m@users.noreply.github.com> * Python: Add tool call/result content types and update connectors and samples (#2971) * Add new AI content types and image tool support Co-authored-by: eavanvalkenburg <13749212+eavanvalkenburg@users.noreply.github.com> * Add Python content types for tool calls/results and image generation tool support Co-authored-by: eavanvalkenburg <13749212+eavanvalkenburg@users.noreply.github.com> * Address review feedback for tool content and samples Co-authored-by: eavanvalkenburg <13749212+eavanvalkenburg@users.noreply.github.com> * Tighten image generation typing and sample tools list Co-authored-by: eavanvalkenburg <13749212+eavanvalkenburg@users.noreply.github.com> * Align image generation output typing Co-authored-by: eavanvalkenburg <13749212+eavanvalkenburg@users.noreply.github.com> * Handle MCP naming, image options mapping, and connector tool content Co-authored-by: eavanvalkenburg <13749212+eavanvalkenburg@users.noreply.github.com> * Allow MCP call in function approval request Co-authored-by: eavanvalkenburg <13749212+eavanvalkenburg@users.noreply.github.com> * Remove raw image_generation tool remapping Co-authored-by: eavanvalkenburg <13749212+eavanvalkenburg@users.noreply.github.com> * Restore Anthropic tool_use to function calls unless code execution Co-authored-by: eavanvalkenburg <13749212+eavanvalkenburg@users.noreply.github.com> * Fix lint issues for hosted file docstring and MCP parsing Co-authored-by: eavanvalkenburg <13749212+eavanvalkenburg@users.noreply.github.com> * Import ChatResponse types in Anthropic client Co-authored-by: eavanvalkenburg <13749212+eavanvalkenburg@users.noreply.github.com> * Fix Anthropics citation type imports and MCP typing for handoff/tools Co-authored-by: eavanvalkenburg <13749212+eavanvalkenburg@users.noreply.github.com> * Skip lightning tests without agentlightning and fix function call import Co-authored-by: eavanvalkenburg <13749212+eavanvalkenburg@users.noreply.github.com> * fix lint on lab package * rebuilt anthropic parsing * redid anthropic parsing * typo * updated parsing and added missing docstrings * fix tests * mypy fixes * second mypy fix * add new class to other samples --------- Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com> Co-authored-by: eavanvalkenburg <13749212+eavanvalkenburg@users.noreply.github.com> Co-authored-by: eavanvalkenburg <github@vanvalkenburg.eu> * Bump Google.GenAI from 0.6.0 to 0.9.0 (#2995) --- updated-dependencies: - dependency-name: Google.GenAI dependency-version: 0.9.0 dependency-type: direct:production update-type: version-update:semver-minor ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: Chris <66376200+crickman@users.noreply.github.com> * Bump js-yaml from 4.1.0 to 4.1.1 in /python/packages/devui/frontend (#3123) Bumps [js-yaml](https://github.com/nodeca/js-yaml) from 4.1.0 to 4.1.1. - [Changelog](https://github.com/nodeca/js-yaml/blob/master/CHANGELOG.md) - [Commits](https://github.com/nodeca/js-yaml/compare/4.1.0...4.1.1) --- updated-dependencies: - dependency-name: js-yaml dependency-version: 4.1.1 dependency-type: indirect ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> * Updated package versions (#3144) * .NET: Bump Microsoft.Agents.AI.OpenAI and Microsoft.Extensions.AI.OpenAI (#2996) * Bump Microsoft.Agents.AI.OpenAI and Microsoft.Extensions.AI.OpenAI Bumps Microsoft.Agents.AI.OpenAI from 1.0.0-preview.251125.1 to 1.0.0-preview.251219.1 Bumps Microsoft.Extensions.AI.OpenAI from 10.1.0-preview.1.25608.1 to 10.1.1-preview.1.25612.2 --- updated-dependencies: - dependency-name: Microsoft.Agents.AI.OpenAI dependency-version: 1.0.0-preview.251219.1 dependency-type: direct:production update-type: version-update:semver-patch - dependency-name: Microsoft.Extensions.AI.OpenAI dependency-version: 10.1.1-preview.1.25612.2 dependency-type: direct:production update-type: version-update:semver-patch - dependency-name: Microsoft.Agents.AI.OpenAI dependency-version: 1.0.0-preview.251219.1 dependency-type: direct:production update-type: version-update:semver-patch - dependency-name: Microsoft.Extensions.AI.OpenAI dependency-version: 10.1.1-preview.1.25612.2 dependency-type: direct:production update-type: version-update:semver-patch ... Signed-off-by: dependabot[bot] <support@github.com> * Fixed samples --------- Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: Chris <66376200+crickman@users.noreply.github.com> Co-authored-by: Mark Wallace <127216156+markwallace-microsoft@users.noreply.github.com> Co-authored-by: Dmytro Struk <13853051+dmytrostruk@users.noreply.github.com> * Python: fix(ag-ui): Execute tools with approval_mode, fix shared state, code cleanup (#3079) * fix(ag-ui): execute tools after approval in human-in-the-loop flow * Fix shared state bug * Bug fix finalized * Refactoring to clean up code * Code cleanup * More fixes * More code cleanup * Add version detection in __init__.py to ruff ignore list * Track agent name with updates for workflow agent (#3146) * Python: Fix AzureAIClient tool call bug for AG-UI use (#3148) * Fiz AzureAIClient tool call bug * Address copilot feedback * Python: multiple bug fixes (#3150) * fix Python: kwargs are not passed to _prepare_thread_and_messages in ChatAgent.run Fixes #3118 * fix Python: [Bug]: model_id versus model_deployment_name is confusing in Azure AI Agents Fixes #3147 * add types * fixed type and docstring * fix(anthropic): fix duplicate ToolCallStartEvent in streaming tool calls (#3051) When processing `input_json_delta` events, the Anthropic client was passing the tool name from the previous `tool_use` event. This caused ag-ui's `_handle_function_call_content` to emit a `ToolCallStartEvent` for every streaming chunk (since it triggers on `if content.name:`). This fix changes the behavior to pass an empty string for `name` in `input_json_delta` events, matching OpenAI's behavior where streaming argument chunks have `name=""`. The initial `tool_use` event still provides the tool name, so only one `ToolCallStartEvent` is emitted. Co-authored-by: Evan Mattson <35585003+moonbox3@users.noreply.github.com> * .NET: [BREAKING] Change GetNewThread and DeserializeThread to async (#3152) * Change GetNewThread and DeserializeThread plus ChatMessageStore and AIContextProvider Factories to async * Merge fixes * Fix Ollama model env var in documentation (#3156) Signed-off-by: Dina Suehiro Jones <dina.s.jones@intel.com> * Python: Add Pydantic request model and OpenAPI tags support to AG-UI FastAPI endpoint (#2522) * feat(ag-ui): Add Pydantic request model and OpenAPI tags support - Add AGUIRequest Pydantic model in _types.py with field descriptions - Update add_agent_framework_fastapi_endpoint() to accept tags parameter - Use AGUIRequest model for automatic validation and OpenAPI schema generation - Export AGUIRequest and DEFAULT_TAGS in __init__.py - Update test_endpoint.py to expect 422 for invalid requests - Add tests for OpenAPI schema, default tags, custom tags, and validation Benefits: - Better API documentation with complete request schema in Swagger UI - Automatic request validation with Pydantic - Organized endpoints under 'AG-UI' tag instead of 'default' - Improved developer experience and type safety Fixes #<issue-number> * test(ag-ui): Add test for internal error handling to achieve 100% coverage - Add test_endpoint_internal_error_handling() to cover exception handling code - Mock copy.deepcopy to simulate internal error during default_state processing - Add type: ignore for FastAPI tags parameter (known pyright compatibility issue) - Achieves 100% test coverage for _endpoint.py (previously missing lines 103-105) * .NET: Improve resolving `AITool` from DI (#3175) * remove localagenttoolregistry * also give the factory method API * Python: Fix MCPStreamableHTTPTool to use new streamable_http_client API (#3088) * Fix MCPStreamableHTTPTool to use new streamable_http_client API with proper httpx client cleanup Co-authored-by: eavanvalkenburg <13749212+eavanvalkenburg@users.noreply.github.com> * Update docstring to reflect new streamable_http_client API usage Co-authored-by: eavanvalkenburg <13749212+eavanvalkenburg@users.noreply.github.com> * Refactor MCPStreamableHTTPTool to accept optional http_client parameter and delegate client creation to streamable_http_client Co-authored-by: eavanvalkenburg <13749212+eavanvalkenburg@users.noreply.github.com> * Update mcp package minimum version to 1.24.0 for streamable_http_client API support Co-authored-by: eavanvalkenburg <13749212+eavanvalkenburg@users.noreply.github.com> * Fix critical bugs: apply headers/timeout/sse_read_timeout when creating httpx client, add version constraint <2, and properly manage client lifecycle Co-authored-by: eavanvalkenburg <13749212+eavanvalkenburg@users.noreply.github.com> * Simplify implementation: remove headers/timeout/sse_read_timeout params, remove kwargs, remove close() override per feedback Co-authored-by: eavanvalkenburg <13749212+eavanvalkenburg@users.noreply.github.com> * Add back **kwargs parameter for backward compatibility (accepted but not used) Co-authored-by: eavanvalkenburg <13749212+eavanvalkenburg@users.noreply.github.com> * Remove unused httpx import from test file Note: The uv.lock file needs to be updated with 'uv sync' to reflect the mcp version constraint change (>=1.24.0,<2) Co-authored-by: eavanvalkenburg <13749212+eavanvalkenburg@users.noreply.github.com> * cicd fixes * udpated samples with headers examples --------- Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com> Co-authored-by: eavanvalkenburg <13749212+eavanvalkenburg@users.noreply.github.com> Co-authored-by: eavanvalkenburg <github@vanvalkenburg.eu> * azureai direct a2a endpoint support (#3127) * Python: [BREAKING]: removed display_name, renamed context_providers, middleware and AggregateContextProvider (#3139) * removed display_name, renamed context_providers, middleware and AggregateContextProvider * fixes * fixed test * testfix * removed mistakenly put back test * updated new test * rename middlewares to middleware * middleware fixes * Python: MCP Improvements: improved connection loss behavior, pagination for loading and a param to control representation (#3154) * pagination support (#2848) added a parse_tool_result param and connection loss (#2884) * fix #3153 * improved connection handling * improved logic * Python: Add declarative workflow runtime (#2815) * Further support for declarative python workflows * Add tests. Clean up for typing and formatting * Improvements and cleanup * Typing cleanup. Improve docstrings * Proper code in docstrings * Fix malformed code-block directive in docstring * Remove dead links * PR feedback * Address PR feedback * Address PR feedback * Remove sl * Update devui frontend * More cleanup * Fix uv lock * Skip Py 3.14 tests as powerfx doesn't support it * Fix mypy error * Fix for tool calls * Removed stale docstring * Fix lint * Standardize on .NET namespaces. Revert DevUI changes (bring in later) * Implement remaining items for Python declarative support to match dotnet * point URL to agent, not to agentcard (#3176) * Python: [BREAKING]: Introducing Options as TypedDict and Generic (#3140) * WIP typeddict for options * updated all clients and ChatAgents * updated everything * added ADR * fix mypy * proper typevar imports * fixed import * fixed other imports * slight update in the sample * updated from feedback * fixes * fixed missing covariants and test fixes * fixed typing * updated anthropic thinking config * ruff fixes * fixed int tests * fix tests and mypy * updated integration tests * updated docstring and test fix * improved options handling in obser * mypy fix * updated a host of integration tests * fix tests * bedrock fix * [BREAKING] Python: Refactor orchestrations (#3023) * Group chat refactoring Part 1; Next: HIL and handoff * Add agent approval flow; next samples * WIP: samples * WIP: HIL samples * Group chat HIL working; next: handoff * Fix group chat tool approval sample * WIP: refactor handoff; next handoff handling * Handoff done; next handoff samples and concurrent and sequential * Handoff samples, concurrent, and sequential done; next Magentic * WIP: magentic; next test with samples + HIL * Magentic Working; next fix all samples and tests * Fix handoff samples; next tests * WIP: fixing tests; some orchestration as agent samples are failing * Group chat unit tests done * Handoff unit tests done * Remove old orchestration_request_info and fix related tests * Magentic unit tests done * Fix samples * Fix test * Fix test 2 * mypy * Address comments * Update readme * Address comments * Address comments 2 * Replace display name * Python: ADR for create/get agent API (#2618) * ADR for create/get agent API * Updated ADR with implementation options * Small updates * Updated decision outcome section * Updated broken links * Small updates * Fixed merge conflicts * Small fix * Updated decision outcome section * Small fixes * Updated provider naming based on client SDK * Add ignored parameter for CodeQL in workflow (#3204) * Implement IReadOnlyList on InMemoryChatMessageStore (#3205) * .NET: Make ChatMessageStore and AIContextProvider context props settable (#3196) * Make ChatMessageStore and AIContextProvider context props setable * Add validation to preserve non-null requirement of certain properties. * Fix broken tests. * Python: Add dependencies param to ag-ui FastAPI endpoint (#3191) * Add dependencies param to ag-ui FastAPI endpoint * Address Copilot feedback * renamed all (#3207) * Python: ADR for simplified get response (#3098) * ADR for simplified get response * updated some language, added agent option and code comparison * small update in sample * added workflows and expanded some points * changed decision and number * updated with stream=False default * .NET: [Breaking] Rename`AgentRunResponse` and `AgentRunResponseUpdate` classes (#3197) * rename AgentRunResponse and AgentRunResponseUpdate classes - part1 * rename varialbles, parameters, methods and tests * rollback unnecessary changes * .NET: [Breaking] Rename AgentRunResponseEvent and AgentRunUpdateEvent classes (#3214) * rename AgentRunResponseEvent and AgentRunUpdateEvent classes * rollback unnecessary changes * Python: Create/Get Agent API for Azure V2 (#3059) * Added get_agent method to Azure AI V2 * Small fixes * Small fix * Removed AzureAIAgentProvider * Added create_agent method * Small fixes * Fixed code interpreter tool mapping * Added agent provider for V2 client * Updated response format handling * Added provider example * Fixed errors * Update python/samples/getting_started/agents/azure_ai/README.md Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Small fix * Updates from merge * Resolved comments * Resolved comments --------- Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Python: Add more specific exceptions to Workflow (#3188) * Add more specifc workflow exceptions * Fix tests * AI comments * Misc * Python: Added AzureAI sample for downloading code interpreter generated files (#3189) * added azure ai code interpreter file download sample * copilot fix suggestions * function name fixes + readme update * small fix * update package versions (#3223) Co-authored-by: Dmytro Struk <13853051+dmytrostruk@users.noreply.github.com> * Python: fix(core): correct FunctionResultContent ordering in WorkflowAgent.merge_updates (#3168) * fix(core): simplify FunctionResultContent ordering in WorkflowAgent.merge_updates * improve comment * Fix name * fix(workflows): rename WorkflowOutputEvent.source_executor_id to executor_id for API consistency (#3166) * Python: fix(ag-ui): add MCP tool support for AG-UI approval flows (#3212) * add MCP tool support for AG-UI approval flows * use attribute in place of property * Python: Properly configure structured outputs based on new options dict (#3213) * Properly configure structured outputs based on new options dict * Fix mypy * .NET: Merge AgentRunOptions.AdditionalProperties into ChatOptions.AdditionalProperties (#3184) * Merge AgentRunOptions.AdditionalProperties into ChatOptions.AdditionalProperties * Fix namespace and typo. * .NET: Update Google.GenAI to 0.11.0 and remove polyfill implementations (#3232) * Initial plan * Update Google.GenAI to 0.11.0 and remove polyfill files Co-authored-by: rogerbarreto <19890735+rogerbarreto@users.noreply.github.com> --------- Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com> Co-authored-by: rogerbarreto <19890735+rogerbarreto@users.noreply.github.com> * .NET: [BREAKING] Renamed CreateAIAgent/GetAIAgent to AsAIAgent (#3222) * Renamed chat client extension method * Additional renaming * Updated documentation * Fixed tests * Small fix * Small fix * Updated DurableAIAgent and fixed integration tests (#3241) * Python: Create/Get Agent API for Azure V1 (#3192) * Added provider implementation for Azure AI V1 * Small fixes * Fixed OpenAPI example * Fixed local MCP example * Fixed hosted MCP example * Fixed file search sample * Small fixes * Resolved comments * Doc updates * Bump azure-core from 1.37.0 to 1.38.0 in /python (#3209) Bumps [azure-core](https://github.com/Azure/azure-sdk-for-python) from 1.37.0 to 1.38.0. - [Release notes](https://github.com/Azure/azure-sdk-for-python/releases) - [Commits](https://github.com/Azure/azure-sdk-for-python/compare/azure-core_1.37.0...azure-core_1.38.0) --- updated-dependencies: - dependency-name: azure-core dependency-version: 1.38.0 dependency-type: indirect ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> * Python: Create/Get Agent API for OpenAI Assistants (#3208) * Added provider implementation * Added example with response format * Small improvements * Python: (AG-UI) Support service-managed thread on AG-UI (#3136) * added service thread support * set service_thread_id to only supplied_thread_id * uses raw_representation to extract the conversation_id * removed accidental edit * updated test to use raw_representation * resolves copilot review feedback * revert back StubAgent, since not used * removed relative module import * removed hasattr check per PR feedback * Create/Get Agent API - fixes and example improvements (#3246) * .NET Purview Middleware: Improve Background Job Runner Injection (#3256) * Clean up background job dependency injection * Fix xml documentation grammar * Python: [BREAKING] Renamed create_agent to as_agent (#3249) * Renamed create_agent to as_agent * Override for as_agent * Added override * Python: Update package version (#3258) * package version 260116 * removed name tags * Python: Fixed Azure chat client for asynchronous filtering (#3260) * Fixed Azure chat client for asynchronous filtering * Updated test * Python: Fixed use_agent_middleware calling private _normalize_messages (#3264) * Fix use_agent_middleware calling private _normalize_messages * Fixed A2A and Copilot Studio agent * Python: Added rai_config to Azure AI agent creation (#3265) * Add kwargs to create_agent method * Added test for kwargs * Addressed comment * Added doc string * Python: Filter conversation_id when passing kwargs to agent as tool (#3266) * Filter conversation_id when passing kwargs to agent as tool * Small fix * Update python/samples/getting_started/agents/azure_ai/README.md Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Update python/samples/getting_started/agents/openai/openai_responses_client_with_agent_as_tool.py Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Update python/samples/getting_started/agents/azure_ai/azure_ai_with_agent_as_tool.py Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Bump actions/setup-dotnet from 5.0.1 to 5.1.0 (#3273) Bumps [actions/setup-dotnet](https://github.com/actions/setup-dotnet) from 5.0.1 to 5.1.0. - [Release notes](https://github.com/actions/setup-dotnet/releases) - [Commits](https://github.com/actions/setup-dotnet/compare/v5.0.1...v5.1.0) --- updated-dependencies: - dependency-name: actions/setup-dotnet dependency-version: 5.1.0 dependency-type: direct:production update-type: version-update:semver-minor ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> * Update ignored checks in merge-gatekeeper workflow * Python: [BREAKING] Make response_format validation errors visible to users (#3274) * Make response_format validation errors visible to users * Small fix * Addressed comments * Python: fix(declarative): Fix MCP tool connection not passed from YAML to Azure AI agent creation API (#3248) * fix(declarative): Fix MCP tool connection not passed from YAML * Add samples to README * Fix mypy * Fix mypy again * Address PR comments * fix #3171, ensure proper form rendering for int (#3201) * Bump uv from 0.9.25 to 0.9.26 in /python (#3288) Bumps [uv](https://github.com/astral-sh/uv) from 0.9.25 to 0.9.26. - [Release notes](https://github.com/astral-sh/uv/releases) - [Changelog](https://github.com/astral-sh/uv/blob/main/CHANGELOG.md) - [Commits](https://github.com/astral-sh/uv/compare/0.9.25...0.9.26) --- updated-dependencies: - dependency-name: uv dependency-version: 0.9.26 dependency-type: direct:development update-type: version-update:semver-patch ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> * Bump ruff from 0.14.11 to 0.14.13 in /python (#3287) Bumps [ruff](https://github.com/astral-sh/ruff) from 0.14.11 to 0.14.13. - [Release notes](https://github.com/astral-sh/ruff/releases) - [Changelog](https://github.com/astral-sh/ruff/blob/main/CHANGELOG.md) - [Commits](https://github.com/astral-sh/ruff/compare/0.14.11...0.14.13) --- updated-dependencies: - dependency-name: ruff dependency-version: 0.14.13 dependency-type: direct:development update-type: version-update:semver-patch ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> * Bump tar from 7.4.3 to 7.5.3 in /python/packages/devui/frontend (#3267) Bumps [tar](https://github.com/isaacs/node-tar) from 7.4.3 to 7.5.3. - [Release notes](https://github.com/isaacs/node-tar/releases) - [Changelog](https://github.com/isaacs/node-tar/blob/main/CHANGELOG.md) - [Commits](https://github.com/isaacs/node-tar/compare/v7.4.3...v7.5.3) --- updated-dependencies: - dependency-name: tar dependency-version: 7.5.3 dependency-type: indirect ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> * .NET: Delete sync extension methods for agent (#3291) * Delete sync extension methods for agent * Fix comments and obsolete attribute * Remove more sync methods. * Fix naming and comments. * Fix unit tests * Python: Fix: Add system_instructions to ChatClient LLM span tracing (#3164) * Fix: Add system_instructions to ChatClient LLM span tracing - Add system_instructions parameter to _capture_messages() calls in _trace_get_response() and _trace_get_streaming_response() - Extract instructions from chat_options in kwargs - Add unit tests to verify system_instructions are captured correctly When using ChatClient with ChatOptions.instructions, the OpenTelemetry LLM span was missing system messages in gen_ai.input.messages and the gen_ai.system_instructions attribute was not being set. This fix aligns the ChatClient-level tracing with the Agent-level tracing which already correctly passes system_instructions. Fixes #3163 Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com> * Add edge case tests for system_instructions - Add test for empty string instructions (should not set attribute) - Add test for list-type instructions (verify multiple items captured) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com> * Simplify: use options.get('instructions') directly instead of kwargs.get('chat_options') Addresses reviewer feedback: - Removed unnecessary chat_options variable from kwargs - Directly access instructions from the options parameter - Updated tests to use dict syntax for options (TypedDict convention) --------- Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com> * Improve PR number handling in workflow (#3302) * Improve PR number handling in workflow Refine PR number extraction and validation method. * Update .github/workflows/python-test-coverage-report.yml Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Fix error message for invalid PR number --------- Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * .NET: Update Microsoft.Extensions.AI.* packages to 10.2.0 (#3211) * Initial plan * Update Microsoft.Extensions.AI.* to 10.2.0 and fix timestamp behavior tests Co-authored-by: stephentoub <2642209+stephentoub@users.noreply.github.com> --------- Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com> Co-authored-by: stephentoub <2642209+stephentoub@users.noreply.github.com> * .NET: Pass AdditionalProperties from parent to child when exposing an agent as a FunctionTool (#3219) * Pass AdditionalProperties from parent to child when exposing an agent as a FunctionTool * Rename variable to improve readability. * Apply suggestions from code review Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Python: [Breaking] Simplified Content types to a single class with classmethod constructors. (#3252) * ported Content to a new model * fixed linting * fixes * fixed data format handling * fix for 3.10 mypy * fix * fix int test * .NET: Durable Agent samples and automated validation for non-Azure Functions (#3042) * Durable Agent samples and automated validation for non-Azure Functions * Update test projects * fix file encoding * Remove AgentThreadMetadata usage * Absorb breaking change from #3152 * Absorb newer breaking changes (AgentRunResponse --> AgentResponse) * Absorb more breaking changes (see #3222) * Improve integration test reliability (isolated task hubs, etc.) * Fix flakey streaming test --------- Co-authored-by: Mark Wallace <127216156+markwallace-microsoft@users.noreply.github.com> * Python: fix(core): handle anyio cancel scope errors during MCP connection cleanup (#3277) * fix(core): handle anyio cancel scope errors during MCP connection cleanup * Address Copilot feedback * Python: fix(ag-ui): properly handle json serialize with handoff workflows as agent (#3275) * fix(ag-ui): properly handle json serialize with handoff workflows as agent * Other improvements around handling non-serializable objects * Bump tomli from 2.3.0 to 2.4.0 in /python (#3182) Bumps [tomli](https://github.com/hukkin/tomli) from 2.3.0 to 2.4.0. - [Changelog](https://github.com/hukkin/tomli/blob/master/CHANGELOG.md) - [Commits](https://github.com/hukkin/tomli/compare/2.3.0...2.4.0) --- updated-dependencies: - dependency-name: tomli dependency-version: 2.4.0 dependency-type: direct:development update-type: version-update:semver-minor ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> * .Net: Fix DebuggerDisplay attribute to reference existing property (#3326) * Initial plan * Fix DebuggerDisplay attribute to use Name instead of DisplayName Co-authored-by: rogerbarreto <19890735+rogerbarreto@users.noreply.github.com> --------- Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com> Co-authored-by: rogerbarreto <19890735+rogerbarreto@users.noreply.github.com> * .NET: Add sample to show multiple AIContextProvider usage (#3284) * Add sample to show multiple AIContextProvider usage * Update comment. * Update messaging in README. * Address PR comments. --------- Co-authored-by: Mark Wallace <127216156+markwallace-microsoft@users.noreply.github.com> * .NET: Update Conversation Sample to use Conversation Id instead (#3180) * Update Conversation Sample to use conversation Id instead * Remove Run infix * Remove the sync GetAIAgent from sample * Python: Fix local MCP tools with `AzureAIProjectAgentProvider` (#3315) * azureai v2 local mcp fix * addressed copilot comments * .NET: Improve readme for agents V2 (#3285) * Improve readme for agents V2 * Architectural justification * Update dotnet/samples/GettingStarted/FoundryAgents/README.md Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Bump pyasn1 from 0.6.1 to 0.6.2 in /python (#3257) Bumps [pyasn1](https://github.com/pyasn1/pyasn1) from 0.6.1 to 0.6.2. - [Release notes](https://github.com/pyasn1/pyasn1/releases) - [Changelog](https://github.com/pyasn1/pyasn1/blob/main/CHANGES.rst) - [Commits](https://github.com/pyasn1/pyasn1/compare/v0.6.1...v0.6.2) --- updated-dependencies: - dependency-name: pyasn1 dependency-version: 0.6.2 dependency-type: indirect ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> * .NET: Fix DebuggerDisplay attribute in AIAgent.cs to reference existing properties (#2985) * Initial plan * Fix DebuggerDisplay attribute in AIAgent.cs to reference existing properties Co-authored-by: stephentoub <2642209+stephentoub@users.noreply.github.com> --------- Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com> Co-authored-by: stephentoub <2642209+stephentoub@users.noreply.github.com> Co-authored-by: Roger Barreto <19890735+rogerbarreto@users.noreply.github.com> * Python: feat(anthropic): Add response_format support for structured outputs (#3301) * fix(anthropic): Add response_format support for structured outputs * only use from options * use native way of response format * ruff lint fix * address comment; handle dict * Updated package versions (#3335) * Set min version of dependent azure-ai-projects to 2.0.0b3 (#3347) * Adding feature collections ADR (#3332) * .NET: [Breaking] Allow passing auth token credential to cosmosdb extensions (#3250) * allow passing token credentials to cosmosdb extensions * Update dotnet/src/Microsoft.Agents.AI.CosmosNoSql/CosmosDBWorkflowExtensions.cs Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Update dotnet/src/Microsoft.Agents.AI.CosmosNoSql/CosmosDBWorkflowExtensions.cs Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Update dotnet/src/Microsoft.Agents.AI.CosmosNoSql/CosmosDBChatExtensions.cs Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * fix: Subworkflows do not work well with HostAsAgent (#3240) Subworkflows run into issues with Checkpointing and the Chat Protocol: * The concurrency rework made subtle changes in behaviour that introduced a hang when using subworkflows with ChatProtocol and streaming execution. * The ResetAsync() implementation in WorkflowHostExecutor was improperly resetting the joinContext - this was happening on restore checkpoint _after_ the join context was attached when * Subworkflows cannot be used as the start node when hosted AsAgent due to inability to treat Catch-All as a Chat Protocol * Subworkflow ownership issue when used in non-concurrent mode after finishing a run Also fixes: * When ChatMessages are output by executors that are not agents, there is no corresponding AgentResponseUpdate/AgentResponse event Breaking Changes * [BREAKING CHANGE] It is possible to provide the wrong RunId when resuming from CheckpointInfo (even though the data already exists on CheckpointInfo) * Python: .NET: Executor source gen for workflow executor routing (#3131) * Roslyn Source Generators for Workflow Executor Routing. * Update dotnet/src/Microsoft.Agents.AI.Workflows.Generators/ExecutorRouteGenerator.cs Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * WIP. * All fixed up except dangling sends/yields attriutes, working on that next. * Add protocol-only generation for SendsMessage/YieldsOutput attributes * Ensuring collections that can change order are sorted to enable pipeline caching. * Improvents per PR feedback. --------- Co-authored-by: alliscode <bentho@microsoft.com> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * .NET: Joslat fix sample issue (#3270) * adds support for labels in edges, fixes rendering of labels in dot and mermaid, adds rendering of labels in edges * Update dotnet/src/Microsoft.Agents.AI.Workflows/Visualization/WorkflowVisualizer.cs Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * escaping edge labels, adding tests for labels containing strange characters that would break the diagram and enabling the previous signature so the API has backwards compatibility. * Unify label in EdgeData * Edge API adjustments, removed useless "sanitizer" * fixed test * Fix in Sample * update --------- Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> Co-authored-by: Jacob Alber <jaalber@microsoft.com> Co-authored-by: Chris <66376200+crickman@users.noreply.github.com> * .NET: Improve unit test coverage for Microsoft.Agents.AI.OpenAI (#3349) * Initial plan * Add unit tests for Microsoft.Agents.AI.OpenAI to improve code coverage Co-authored-by: rogerbarreto <19890735+rogerbarreto@users.noreply.github.com> * Address code review feedback: remove unused using directives Co-authored-by: rogerbarreto <19890735+rogerbarreto@users.noreply.github.com> * Fix format issues: file encoding and remove unused using directives Co-authored-by: rogerbarreto <19890735+rogerbarreto@users.noreply.github.com> * Fix redundant cast error by using named parameter Co-authored-by: rogerbarreto <19890735+rogerbarreto@users.noreply.github.com> * Remove excessive inline comments per PR review feedback --------- Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com> Co-authored-by: rogerbarreto <19890735+rogerbarreto@users.noreply.github.com> * Revert to main * Python: Fix: Verify types during checkpoint deserialization to prevent marker spoofing (#3243) * Initial plan * Add validation for reserved keywords in checkpoint encoding/decoding Co-authored-by: TaoChenOSU <12570346+TaoChenOSU@users.noreply.github.com> * Refactor to eliminate duplicate code in model protocol detection Co-authored-by: TaoChenOSU <12570346+TaoChenOSU@users.noreply.github.com> * Fix pyright type narrowing issue for dataclass check Co-authored-by: TaoChenOSU <12570346+TaoChenOSU@users.noreply.github.com> * Add comprehensive unit tests for checkpoint encoding Co-authored-by: TaoChenOSU <12570346+TaoChenOSU@users.noreply.github.com> * Remove serialization-time reserved keyword validation to fix failing tests The serialization-time validation was too aggressive and blocked legitimate use cases where encoded data was being re-encoded. Security is now enforced only at deserialization time by validating that classes marked with DATACLASS_MARKER are actual dataclasses and classes marked with MODEL_MARKER actually support the model protocol. Co-authored-by: TaoChenOSU <12570346+TaoChenOSU@users.noreply.github.com> * Apply ruff formatting to checkpoint encoding file Co-authored-by: TaoChenOSU <12570346+TaoChenOSU@users.noreply.github.com> * Changes before error encountered Co-authored-by: TaoChenOSU <12570346+TaoChenOSU@users.noreply.github.com> * Revert "Changes before error encountered" This reverts commit f515b880dc. --------- Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com> Co-authored-by: TaoChenOSU <12570346+TaoChenOSU@users.noreply.github.com> Co-authored-by: Tao Chen <taochen@microsoft.com> * Python: Fix azurefunctions MCP tool invocation to use correct agent (#3339) * MCP tool fix for azurefunctions * Moving logic to check for thread id * Adding ReflectExecutors method to Workflow. (#3389) * Fix merge conflicts * Python: [BREAKING] simplify ag-ui run logic, fix mcp bugs, fix anthropic client issues in ag-ui (#3322) * Refactor ag-ui to simplify flow * Refactoring * Fix backend tool * Update tests * Improvements * Fix mypy * Fixes * Fix json serialize errors * Python: fix(core): filter out internal args when passing kwargs to MCP tools (#3292) * fix(core): filter conversation_id when passing kwargs to MCP tools * Filter out options too * Fix uv.lock conflict * Python: Added tests for OpenAI content types + Unit test improvement (#3259) * added tests for content types+ unit test improvement * small fixes * small fix * Python: Prefer runtime `kwargs` for `conversation_id` in OpenAI Responses client (#3312) * prefer kwargs conversation_id over options * addressed comments * Python: Azure AI mapping HostedImageGenerationTool to ImageGenTool (#3263) * azureai image gen sample fix * mypy fixes * addressed comments + mapping updates * image model fix * content type fix * Python: add(azure-ai): support reasoning config for AzureAIClient (#3403) * add(azure-ai): support reasoning config for AzureAIClient * Update sample * Merge main * improvements * improve sample * .NET: Allow overriding the ChatMessageStore to be used per agent run. (#3330) * Allow overriding the ChatMessageStore to be used per agent run. * Fix typos * Fix Add and add TryAdd, Contains and Remove * Update instructions to require automatically building and formatting (#3412) * .NET: Rename ChatMessageStore to ChatHistoryProvider (#3375) * Rename ChatMessageStore to ChatHistoryProvider * Fix merge issue * Fixed PR comments * Fix tests after property rename * Add unit tests and fix merge issues * Fix encoding --------- Signed-off-by: dependabot[bot] <support@github.com> Signed-off-by: Dina Suehiro Jones <dina.s.jones@intel.com> Co-authored-by: Tao Chen <taochen@microsoft.com> Co-authored-by: Kurt <65111699+q33566@users.noreply.github.com> Co-authored-by: Evan Mattson <35585003+moonbox3@users.noreply.github.com> Co-authored-by: SergeyMenshykh <68852919+SergeyMenshykh@users.noreply.github.com> Co-authored-by: Korolev Dmitry <deagle.gross@gmail.com> Co-authored-by: Mark Wallace <127216156+markwallace-microsoft@users.noreply.github.com> Co-authored-by: Copilot <198982749+Copilot@users.noreply.github.com> Co-authored-by: rogerbarreto <19890735+rogerbarreto@users.noreply.github.com> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: Eduard van Valkenburg <eavanvalkenburg@users.noreply.github.com> Co-authored-by: Dmytro Struk <13853051+dmytrostruk@users.noreply.github.com> Co-authored-by: Chris <66376200+crickman@users.noreply.github.com> Co-authored-by: Jose Luis Latorre Millas <joslat@gmail.com> Co-authored-by: Jacob Alber <jaalber@microsoft.com> Co-authored-by: Richard Ortega <richardjortega@gmail.com> Co-authored-by: 刘邦学AI <lbbniu@gmail.com> Co-authored-by: Stephen Toub <stoub@microsoft.com> Co-authored-by: Nico Möller <nkm-moeller@mail.de> Co-authored-by: Chris Gillum <cgillum@microsoft.com> Co-authored-by: Giles Odigwe <79032838+giles17@users.noreply.github.com> Co-authored-by: Phillip Hoff <phillip.hoff@gmail.com> Co-authored-by: Ege Ozan Özyedek <36128615+egeozanozyedek@users.noreply.github.com> Co-authored-by: samueljohnsiby <66901393+samueljohnsiby@users.noreply.github.com> Co-authored-by: Evan Mattson <evan.mattson@microsoft.com> Co-authored-by: Hao Luo <338265+howlowck@users.noreply.github.com> Co-authored-by: Victor Dibia <chuvidi2003@gmail.com> Co-authored-by: stephentoub <2642209+stephentoub@users.noreply.github.com> Co-authored-by: Jacob Viau <javia@microsoft.com> Co-authored-by: SuperKenVery <39673849+SuperKenVery@users.noreply.github.com> Co-authored-by: Sunil Dutta <dutta.2003@gmail.com> Co-authored-by: Sunil Dutta <sunil.dutta@penske.com> Co-authored-by: budgetboardingai <apurva.sharma31@gmail.com> Co-authored-by: Syrine Chelly <62653967+SyChell@users.noreply.github.com> Co-authored-by: SergeyMenshykh <sergemenshikh@gmail.com> Co-authored-by: westey <164392973+westey-m@users.noreply.github.com> Co-authored-by: takanori-terai <123897708+takanori-terai@users.noreply.github.com> Co-authored-by: claude89757 <138977524+claude89757@users.noreply.github.com> Co-authored-by: Gavin Aguiar <80794152+gavin-aguiar@users.noreply.github.com> Co-authored-by: Sukeesh <vsukeeshbabu@gmail.com> Co-authored-by: eavanvalkenburg <13749212+eavanvalkenburg@users.noreply.github.com> Co-authored-by: eavanvalkenburg <github@vanvalkenburg.eu> Co-authored-by: Ao Chen <chenao3220@gmail.com> Co-authored-by: Dina Suehiro Jones <dina.s.jones@intel.com> Co-authored-by: eoindoherty1 <eoindoherty@microsoft.com> Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com> Co-authored-by: Darren Cohen <39422044+dargilco@users.noreply.github.com> Co-authored-by: Ben Thomas <ben.thomas@microsoft.com> Co-authored-by: alliscode <bentho@microsoft.com> Co-authored-by: TaoChenOSU <12570346+TaoChenOSU@users.noreply.github.com> Co-authored-by: Shyju Krishnankutty <connectshyju@gmail.com>
2026-06-16 21:04:09 +08:00 · 2026-01-23 10:31:54 -08:00
parent ff839435a2
commit 172423aab7
552 changed files with 23634 additions and 14420 deletions
@@ -6,13 +6,6 @@ import importlib.metadata

 from ._agent import AgentFrameworkAgent
 from ._client import AGUIChatClient
-from ._confirmation_strategies import (
-    ConfirmationStrategy,
-    DefaultConfirmationStrategy,
-    DocumentWriterConfirmationStrategy,
-    RecipeConfirmationStrategy,
-    TaskPlannerConfirmationStrategy,
-)
 from ._endpoint import add_agent_framework_fastapi_endpoint
 from ._event_converters import AGUIEventConverter
 from ._http_service import AGUIHttpService
@@ -35,13 +28,8 @@ __all__ = [
    "AGUIHttpService",
    "AGUIRequest",
    "AgentState",
-    "ConfirmationStrategy",
-    "DefaultConfirmationStrategy",
    "PredictStateConfig",
    "RunMetadata",
-    "TaskPlannerConfirmationStrategy",
-    "RecipeConfirmationStrategy",
-    "DocumentWriterConfirmationStrategy",
    "DEFAULT_TAGS",
    "__version__",
 ]
@@ -1,6 +1,6 @@
 # Copyright (c) Microsoft. All rights reserved.

-"""AgentFrameworkAgent wrapper for AG-UI protocol - Clean Architecture."""
+"""AgentFrameworkAgent wrapper for AG-UI protocol."""

 from collections.abc import AsyncGenerator
 from typing import Any, cast
@@ -8,13 +8,7 @@ from typing import Any, cast
 from ag_ui.core import BaseEvent
 from agent_framework import AgentProtocol

-from ._confirmation_strategies import ConfirmationStrategy, DefaultConfirmationStrategy
-from ._orchestrators import (
-    DefaultOrchestrator,
-    ExecutionContext,
-    HumanInTheLoopOrchestrator,
-    Orchestrator,
-)
+from ._run import run_agent_stream


 class AgentConfig:
@@ -33,7 +27,7 @@ class AgentConfig:
            state_schema: Optional state schema for state management; accepts dict or Pydantic model/class
            predict_state_config: Configuration for predictive state updates
            use_service_thread: Whether the agent thread is service-managed
-            require_confirmation: Whether predictive updates require confirmation
+            require_confirmation: Whether predictive updates require user confirmation before applying
        """
        self.state_schema = self._normalize_state_schema(state_schema)
        self.predict_state_config = predict_state_config or {}
@@ -58,12 +52,12 @@ class AgentConfig:
            base_model_type = None

        if base_model_type is not None and isinstance(state_schema, base_model_type):
-            schema_dict = state_schema.__class__.model_json_schema()
+            schema_dict = state_schema.__class__.model_json_schema()  # type: ignore[union-attr]
            return schema_dict.get("properties", {}) or {}

        if base_model_type is not None and isinstance(state_schema, type) and issubclass(state_schema, base_model_type):
-            schema_dict = state_schema.model_json_schema()
-            return schema_dict.get("properties", {}) or {}
+            schema_dict = state_schema.model_json_schema()  # type: ignore[union-attr]
+            return schema_dict.get("properties", {}) or {}  # type: ignore

        return {}

@@ -72,12 +66,7 @@ class AgentFrameworkAgent:
    """Wraps Agent Framework agents for AG-UI protocol compatibility.

    Translates between Agent Framework's AgentProtocol and AG-UI's event-based
-    protocol. Uses orchestrators to handle different execution flows (standard
-    execution, human-in-the-loop, etc.). Orchestrators are checked in order;
-    the first matching orchestrator handles the request.
-
-    Supports predictive state updates for agentic generative UI, with optional
-    confirmation requirements configurable per use case.
+    protocol. Follows a simple linear flow: RunStarted -> content events -> RunFinished.
    """

    def __init__(
@@ -88,9 +77,7 @@ class AgentFrameworkAgent:
        state_schema: Any | None = None,
        predict_state_config: dict[str, dict[str, str]] | None = None,
        require_confirmation: bool = True,
-        orchestrators: list[Orchestrator] | None = None,
        use_service_thread: bool = False,
-        confirmation_strategy: ConfirmationStrategy | None = None,
    ):
        """Initialize the AG-UI compatible agent wrapper.

@@ -99,15 +86,9 @@ class AgentFrameworkAgent:
            name: Optional name for the agent
            description: Optional description
            state_schema: Optional state schema for state management; accepts dict or Pydantic model/class
-            predict_state_config: Configuration for predictive state updates.
-                Format: {"state_key": {"tool": "tool_name", "tool_argument": "arg_name"}}
-            require_confirmation: Whether predictive updates require confirmation.
-                Set to False for agentic generative UI that updates automatically.
-            orchestrators: Custom orchestrators (auto-configured if None).
-                Orchestrators are checked in order; first match handles the request.
-            use_service_thread: Whether the agent thread is service-managed.
-            confirmation_strategy: Strategy for generating confirmation messages.
-                Defaults to DefaultConfirmationStrategy if None.
+            predict_state_config: Configuration for predictive state updates
+            require_confirmation: Whether predictive updates require user confirmation before applying
+            use_service_thread: Whether the agent thread is service-managed
        """
        self.agent = agent
        self.name = name or getattr(agent, "name", "agent")
@@ -120,74 +101,17 @@ class AgentFrameworkAgent:
            require_confirmation=require_confirmation,
        )

-        # Configure orchestrators
-        if orchestrators is None:
-            self.orchestrators = self._default_orchestrators()
-        else:
-            self.orchestrators = orchestrators
-
-        # Configure confirmation strategy
-        if confirmation_strategy is None:
-            self.confirmation_strategy: ConfirmationStrategy = DefaultConfirmationStrategy()
-        else:
-            self.confirmation_strategy = confirmation_strategy
-
-    def _default_orchestrators(self) -> list[Orchestrator]:
-        """Create default orchestrator chain.
-
-        Returns:
-            List of orchestrators in priority order. First matching orchestrator
-            handles the request, so order matters.
-        """
-        return [
-            HumanInTheLoopOrchestrator(),  # Handle tool approval responses
-            # Add more specialized orchestrators here as needed
-            DefaultOrchestrator(),  # Fallback: standard agent execution
-        ]
-
    async def run_agent(
        self,
        input_data: dict[str, Any],
    ) -> AsyncGenerator[BaseEvent, None]:
        """Run the agent and yield AG-UI events.

-        This is the ONLY public method - much simpler than the original 376-line
-        implementation. All orchestration logic has been extracted into dedicated
-        Orchestrator classes.
-
-        The method creates an ExecutionContext with all needed data, then finds
-        the first orchestrator that can handle the request and delegates to it.
-
        Args:
            input_data: The AG-UI run input containing messages, state, etc.

        Yields:
            AG-UI events
-
-        Raises:
-            RuntimeError: If no orchestrator matches (should never happen if
-                DefaultOrchestrator is last in the chain)
        """
-        # Create execution context with all needed data
-        context = ExecutionContext(
-            input_data=input_data,
-            agent=self.agent,
-            config=self.config,
-            confirmation_strategy=self.confirmation_strategy,
-        )
-
-        # Find matching orchestrator and execute
-        for orchestrator in self.orchestrators:
-            if orchestrator.can_handle(context):
-                async for event in orchestrator.run(context):
-                    yield event
-                return
-
-        # Should never reach here if DefaultOrchestrator is last
-        raise RuntimeError("No orchestrator matched - check configuration")
-
-
-__all__ = [
-    "AgentFrameworkAgent",
-    "AgentConfig",
-]
+        async for event in run_agent_stream(input_data, self.agent, self.config):
+            yield event
@@ -17,12 +17,10 @@ from agent_framework import (
    ChatMessage,
    ChatResponse,
    ChatResponseUpdate,
-    DataContent,
-    FunctionCallContent,
+    Content,
+    use_chat_middleware,
+    use_function_invocation,
 )
-from agent_framework._middleware import use_chat_middleware
-from agent_framework._tools import use_function_invocation
-from agent_framework._types import BaseContent, Contents
 from agent_framework.observability import use_instrumentation

 from ._event_converters import AGUIEventConverter
@@ -53,26 +51,11 @@ else:
 logger: logging.Logger = logging.getLogger(__name__)


-class ServerFunctionCallContent(BaseContent):
-    """Wrapper for server function calls to prevent client re-execution.
-
-    All function calls from the remote server are server-side executions.
-    This wrapper prevents @use_function_invocation from trying to execute them again.
-    """
-
-    function_call_content: FunctionCallContent
-
-    def __init__(self, function_call_content: FunctionCallContent) -> None:
-        """Initialize with the function call content."""
-        super().__init__(type="server_function_call")
-        self.function_call_content = function_call_content
-
-
-def _unwrap_server_function_call_contents(contents: MutableSequence[Contents | dict[str, Any]]) -> None:
-    """Replace ServerFunctionCallContent instances with their underlying call content."""
+def _unwrap_server_function_call_contents(contents: MutableSequence[Content | dict[str, Any]]) -> None:
+    """Replace server_function_call instances with their underlying call content."""
    for idx, content in enumerate(contents):
-        if isinstance(content, ServerFunctionCallContent):
-            contents[idx] = content.function_call_content  # type: ignore[assignment]
+        if content.type == "server_function_call":  # type: ignore[union-attr]
+            contents[idx] = content.function_call  # type: ignore[assignment, union-attr]


 TBaseChatClient = TypeVar("TBaseChatClient", bound=type[BaseChatClient[Any]])
@@ -91,9 +74,9 @@ def _apply_server_function_call_unwrap(chat_client: TBaseChatClient) -> TBaseCha
    original_get_streaming_response = chat_client.get_streaming_response

    @wraps(original_get_streaming_response)
-    async def streaming_wrapper(self, *args: Any, **kwargs: Any) -> AsyncIterable[ChatResponseUpdate]:
+    async def streaming_wrapper(self: Any, *args: Any, **kwargs: Any) -> AsyncIterable[ChatResponseUpdate]:
        async for update in original_get_streaming_response(self, *args, **kwargs):
-            _unwrap_server_function_call_contents(cast(MutableSequence[Contents | dict[str, Any]], update.contents))
+            _unwrap_server_function_call_contents(cast(MutableSequence[Content | dict[str, Any]], update.contents))
            yield update

    chat_client.get_streaming_response = streaming_wrapper  # type: ignore[assignment]
@@ -101,13 +84,11 @@ def _apply_server_function_call_unwrap(chat_client: TBaseChatClient) -> TBaseCha
    original_get_response = chat_client.get_response

    @wraps(original_get_response)
-    async def response_wrapper(self, *args: Any, **kwargs: Any) -> ChatResponse:
+    async def response_wrapper(self: Any, *args: Any, **kwargs: Any) -> ChatResponse:
        response = await original_get_response(self, *args, **kwargs)
        if response.messages:
            for message in response.messages:
-                _unwrap_server_function_call_contents(
-                    cast(MutableSequence[Contents | dict[str, Any]], message.contents)
-                )
+                _unwrap_server_function_call_contents(cast(MutableSequence[Content | dict[str, Any]], message.contents))
        return response

    chat_client.get_response = response_wrapper  # type: ignore[assignment]
@@ -289,13 +270,13 @@ class AGUIChatClient(BaseChatClient[TAGUIChatOptions], Generic[TAGUIChatOptions]
        last_message = messages[-1]

        for content in last_message.contents:
-            if isinstance(content, DataContent) and content.media_type == "application/json":
+            if isinstance(content, Content) and content.type == "data" and content.media_type == "application/json":
                try:
                    uri = content.uri
-                    if uri.startswith("data:application/json;base64,"):
+                    if uri.startswith("data:application/json;base64,"):  # type: ignore[union-attr]
                        import base64

-                        encoded_data = uri.split(",", 1)[1]
+                        encoded_data = uri.split(",", 1)[1]  # type: ignore[union-attr]
                        decoded_bytes = base64.b64decode(encoded_data)
                        state = json.loads(decoded_bytes.decode("utf-8"))

@@ -433,19 +414,19 @@ class AGUIChatClient(BaseChatClient[TAGUIChatOptions], Generic[TAGUIChatOptions]
                )
                # Distinguish client vs server tools
                for i, content in enumerate(update.contents):
-                    if isinstance(content, FunctionCallContent):
+                    if content.type == "function_call":  # type: ignore[attr-defined]
                        logger.debug(
-                            f"[AGUIChatClient] Function call: {content.name}, in client_tool_set: {content.name in client_tool_set}"
+                            f"[AGUIChatClient] Function call: {content.name}, in client_tool_set: {content.name in client_tool_set}"  # type: ignore[attr-defined]
                        )
-                        if content.name in client_tool_set:
+                        if content.name in client_tool_set:  # type: ignore[attr-defined]
                            # Client tool - let @use_function_invocation execute it
-                            if not content.additional_properties:
-                                content.additional_properties = {}
-                            content.additional_properties["agui_thread_id"] = thread_id
+                            if not content.additional_properties:  # type: ignore[attr-defined]
+                                content.additional_properties = {}  # type: ignore[attr-defined]
+                            content.additional_properties["agui_thread_id"] = thread_id  # type: ignore[attr-defined]
                        else:
                            # Server tool - wrap so @use_function_invocation ignores it
-                            logger.debug(f"[AGUIChatClient] Wrapping server tool: {content.name}")
-                            self._register_server_tool_placeholder(content.name)
-                            update.contents[i] = ServerFunctionCallContent(content)  # type: ignore
+                            logger.debug(f"[AGUIChatClient] Wrapping server tool: {content.name}")  # type: ignore[union-attr]
+                            self._register_server_tool_placeholder(content.name)  # type: ignore[arg-type]
+                            update.contents[i] = Content(type="server_function_call", function_call=content)  # type: ignore

                yield update
@@ -1,217 +0,0 @@
-# Copyright (c) Microsoft. All rights reserved.
-
-"""Confirmation strategies for human-in-the-loop approval flows.
-
-Each agent can provide a custom confirmation strategy to generate domain-specific
-messages when users approve or reject changes/actions.
-"""
-
-from abc import ABC, abstractmethod
-from typing import Any
-
-
-class ConfirmationStrategy(ABC):
-    """Strategy for generating confirmation messages during human-in-the-loop flows.
-
-    Subclasses must define the message properties. The methods use those properties
-    by default, but can be overridden for complete customization.
-    """
-
-    @property
-    @abstractmethod
-    def approval_header(self) -> str:
-        """Header for approval accepted message. Must be overridden."""
-        ...
-
-    @property
-    @abstractmethod
-    def approval_footer(self) -> str:
-        """Footer for approval accepted message. Must be overridden."""
-        ...
-
-    @property
-    @abstractmethod
-    def rejection_message(self) -> str:
-        """Message when user rejects. Must be overridden."""
-        ...
-
-    @property
-    @abstractmethod
-    def state_confirmed_message(self) -> str:
-        """Message when state is confirmed. Must be overridden."""
-        ...
-
-    @property
-    @abstractmethod
-    def state_rejected_message(self) -> str:
-        """Message when state is rejected. Must be overridden."""
-        ...
-
-    def on_approval_accepted(self, steps: list[dict[str, Any]]) -> str:
-        """Generate message when user approves function execution.
-
-        Default implementation uses header/footer properties.
-        Override for complete customization.
-
-        Args:
-            steps: List of approved steps with 'description', 'status', etc.
-
-        Returns:
-            Message to display to user
-        """
-        enabled_steps = [s for s in steps if s.get("status") == "enabled"]
-        message_parts = [self.approval_header.format(count=len(enabled_steps))]
-        for i, step in enumerate(enabled_steps, 1):
-            message_parts.append(f"{i}. {step['description']}\n")
-        message_parts.append(self.approval_footer)
-        return "".join(message_parts)
-
-    def on_approval_rejected(self, steps: list[dict[str, Any]]) -> str:
-        """Generate message when user rejects function execution.
-
-        Args:
-            steps: List of rejected steps
-
-        Returns:
-            Message to display to user
-        """
-        return self.rejection_message
-
-    def on_state_confirmed(self) -> str:
-        """Generate message when user confirms predictive state changes.
-
-        Returns:
-            Message to display to user
-        """
-        return self.state_confirmed_message
-
-    def on_state_rejected(self) -> str:
-        """Generate message when user rejects predictive state changes.
-
-        Returns:
-            Message to display to user
-        """
-        return self.state_rejected_message
-
-
-class DefaultConfirmationStrategy(ConfirmationStrategy):
-    """Generic confirmation messages suitable for most agents."""
-
-    @property
-    def approval_header(self) -> str:
-        return "Executing {count} approved steps:\n\n"
-
-    @property
-    def approval_footer(self) -> str:
-        return "\nAll steps completed successfully!"
-
-    @property
-    def rejection_message(self) -> str:
-        return "No problem! What would you like me to change about the plan?"
-
-    @property
-    def state_confirmed_message(self) -> str:
-        return "Changes confirmed and applied successfully!"
-
-    @property
-    def state_rejected_message(self) -> str:
-        return "No problem! What would you like me to change?"
-
-
-class TaskPlannerConfirmationStrategy(ConfirmationStrategy):
-    """Domain-specific confirmation messages for task planning agents."""
-
-    @property
-    def approval_header(self) -> str:
-        return "Executing your requested tasks:\n\n"
-
-    @property
-    def approval_footer(self) -> str:
-        return "\nAll tasks completed successfully!"
-
-    @property
-    def rejection_message(self) -> str:
-        return "No problem! Let me revise the plan. What would you like me to change?"
-
-    @property
-    def state_confirmed_message(self) -> str:
-        return "Tasks confirmed and ready to execute!"
-
-    @property
-    def state_rejected_message(self) -> str:
-        return "No problem! How should I adjust the task list?"
-
-
-class RecipeConfirmationStrategy(ConfirmationStrategy):
-    """Domain-specific confirmation messages for recipe agents."""
-
-    @property
-    def approval_header(self) -> str:
-        return "Updating your recipe:\n\n"
-
-    @property
-    def approval_footer(self) -> str:
-        return "\nRecipe updated successfully!"
-
-    @property
-    def rejection_message(self) -> str:
-        return "No problem! What ingredients or steps should I change?"
-
-    @property
-    def state_confirmed_message(self) -> str:
-        return "Recipe changes applied successfully!"
-
-    @property
-    def state_rejected_message(self) -> str:
-        return "No problem! What would you like me to adjust in the recipe?"
-
-
-class DocumentWriterConfirmationStrategy(ConfirmationStrategy):
-    """Domain-specific confirmation messages for document writing agents."""
-
-    @property
-    def approval_header(self) -> str:
-        return "Applying your edits:\n\n"
-
-    @property
-    def approval_footer(self) -> str:
-        return "\nDocument updated successfully!"
-
-    @property
-    def rejection_message(self) -> str:
-        return "No problem! Which changes should I keep or modify?"
-
-    @property
-    def state_confirmed_message(self) -> str:
-        return "Document edits applied!"
-
-    @property
-    def state_rejected_message(self) -> str:
-        return "No problem! What should I change about the document?"
-
-
-def apply_confirmation_strategy(
-    strategy: ConfirmationStrategy | None,
-    accepted: bool,
-    steps: list[dict[str, Any]],
-) -> str:
-    """Apply a confirmation strategy to generate a message.
-
-    This helper consolidates the pattern used in multiple orchestrators.
-
-    Args:
-        strategy: Strategy to use, or None for default
-        accepted: Whether the user approved
-        steps: List of steps (may be empty for state confirmations)
-
-    Returns:
-        Generated message string
-    """
-    if strategy is None:
-        strategy = DefaultConfirmationStrategy()
-
-    if not steps:
-        # State confirmation (no steps)
-        return strategy.on_state_confirmed() if accepted else strategy.on_state_rejected()
-    # Step-based approval
-    return strategy.on_approval_accepted(steps) if accepted else strategy.on_approval_rejected(steps)
@@ -4,7 +4,7 @@

 import copy
 import logging
-from collections.abc import Sequence
+from collections.abc import AsyncGenerator, Sequence
 from typing import Any

 from ag_ui.encoder import EventEncoder
@@ -56,8 +56,8 @@ def add_agent_framework_fastapi_endpoint(
    else:
        wrapped_agent = agent

-    @app.post(path, tags=tags or ["AG-UI"], dependencies=dependencies)  # type: ignore[arg-type]
-    async def agent_endpoint(request_body: AGUIRequest):  # type: ignore[misc]
+    @app.post(path, tags=tags or ["AG-UI"], dependencies=dependencies, response_model=None)  # type: ignore[arg-type]
+    async def agent_endpoint(request_body: AGUIRequest) -> StreamingResponse | dict[str, str]:
        """Handle AG-UI agent requests.

        Note: Function is accessed via FastAPI's decorator registration,
@@ -77,17 +77,19 @@ def add_agent_framework_fastapi_endpoint(
            )
            logger.info(f"Received request at {path}: {input_data.get('run_id', 'no-run-id')}")

-            async def event_generator():
+            async def event_generator() -> AsyncGenerator[str, None]:
                encoder = EventEncoder()
                event_count = 0
                async for event in wrapped_agent.run_agent(input_data):
                    event_count += 1
-                    logger.debug(f"[{path}] Event {event_count}: {type(event).__name__}")
-
-                    # Log event payload for debugging
-                    if hasattr(event, "model_dump"):
-                        event_data = event.model_dump(exclude_none=True)
-                        logger.debug(f"[{path}] Event payload: {event_data}")
+                    event_type_name = getattr(event, "type", type(event).__name__)
+                    # Log important events at INFO level
+                    if "TOOL_CALL" in str(event_type_name) or "RUN" in str(event_type_name):
+                        if hasattr(event, "model_dump"):
+                            event_data = event.model_dump(exclude_none=True)
+                            logger.info(f"[{path}] Event {event_count}: {event_type_name} - {event_data}")
+                        else:
+                            logger.info(f"[{path}] Event {event_count}: {event_type_name}")

                    encoded = encoder.encode(event)
                    logger.debug(
@@ -6,12 +6,9 @@ from typing import Any

 from agent_framework import (
    ChatResponseUpdate,
-    ErrorContent,
+    Content,
    FinishReason,
-    FunctionCallContent,
-    FunctionResultContent,
    Role,
-    TextContent,
 )


@@ -117,7 +114,7 @@ class AGUIEventConverter:
        return ChatResponseUpdate(
            role=Role.ASSISTANT,
            message_id=self.current_message_id,
-            contents=[TextContent(text=delta)],
+            contents=[Content.from_text(text=delta)],
        )

    def _handle_text_message_end(self, event: dict[str, Any]) -> ChatResponseUpdate | None:
@@ -133,7 +130,7 @@ class AGUIEventConverter:
        return ChatResponseUpdate(
            role=Role.ASSISTANT,
            contents=[
-                FunctionCallContent(
+                Content.from_function_call(
                    call_id=self.current_tool_call_id or "",
                    name=self.current_tool_name or "",
                    arguments="",
@@ -149,7 +146,7 @@ class AGUIEventConverter:
        return ChatResponseUpdate(
            role=Role.ASSISTANT,
            contents=[
-                FunctionCallContent(
+                Content.from_function_call(
                    call_id=self.current_tool_call_id or "",
                    name=self.current_tool_name or "",
                    arguments=delta,
@@ -170,7 +167,7 @@ class AGUIEventConverter:
        return ChatResponseUpdate(
            role=Role.TOOL,
            contents=[
-                FunctionResultContent(
+                Content.from_function_result(
                    call_id=tool_call_id,
                    result=result,
                )
@@ -197,7 +194,7 @@ class AGUIEventConverter:
            role=Role.ASSISTANT,
            finish_reason=FinishReason.CONTENT_FILTER,
            contents=[
-                ErrorContent(
+                Content.from_error(
                    message=error_message,
                    error_code="RUN_ERROR",
                )
@@ -1,586 +0,0 @@
-# Copyright (c) Microsoft. All rights reserved.
-
-"""Event bridge for converting Agent Framework events to AG-UI protocol."""
-
-import json
-import logging
-import re
-from copy import deepcopy
-from typing import Any
-
-from ag_ui.core import (
-    BaseEvent,
-    CustomEvent,
-    RunFinishedEvent,
-    RunStartedEvent,
-    StateDeltaEvent,
-    StateSnapshotEvent,
-    TextMessageContentEvent,
-    TextMessageEndEvent,
-    TextMessageStartEvent,
-    ToolCallArgsEvent,
-    ToolCallEndEvent,
-    ToolCallResultEvent,
-    ToolCallStartEvent,
-)
-from agent_framework import (
-    AgentResponseUpdate,
-    FunctionApprovalRequestContent,
-    FunctionCallContent,
-    FunctionResultContent,
-    TextContent,
-    prepare_function_call_results,
-)
-
-from ._utils import extract_state_from_tool_args, generate_event_id, safe_json_parse
-
-logger = logging.getLogger(__name__)
-
-
-class AgentFrameworkEventBridge:
-    """Converts Agent Framework responses to AG-UI events."""
-
-    def __init__(
-        self,
-        run_id: str,
-        thread_id: str,
-        predict_state_config: dict[str, dict[str, str]] | None = None,
-        current_state: dict[str, Any] | None = None,
-        skip_text_content: bool = False,
-        require_confirmation: bool = True,
-        approval_tool_name: str | None = None,
-    ) -> None:
-        """
-        Initialize the event bridge.
-
-        Args:
-            run_id: The run identifier.
-            thread_id: The thread identifier.
-            predict_state_config: Configuration for predictive state updates.
-                Format: {"state_key": {"tool": "tool_name", "tool_argument": "arg_name"}}
-            current_state: Reference to the current state dict for tracking updates.
-            skip_text_content: If True, skip emitting TextMessageContentEvents (for structured outputs).
-            require_confirmation: Whether predictive state updates require user confirmation.
-        """
-        self.run_id = run_id
-        self.thread_id = thread_id
-        self.current_message_id: str | None = None
-        self.current_tool_call_id: str | None = None
-        self.current_tool_call_name: str | None = None  # Track the tool name across streaming chunks
-        self.predict_state_config = predict_state_config or {}
-        self.current_state = current_state or {}
-        self.pending_state_updates: dict[str, Any] = {}  # Track updates from tool calls
-        self.skip_text_content = skip_text_content
-        self.require_confirmation = require_confirmation
-        self.approval_tool_name = approval_tool_name
-
-        # For predictive state updates: accumulate streaming arguments
-        self.streaming_tool_args: str = ""  # Accumulated JSON string
-        self.last_emitted_state: dict[str, Any] = {}  # Track last emitted state to avoid duplicates
-        self.state_delta_count: int = 0  # Counter for sampling log output
-        self.should_stop_after_confirm: bool = False  # Flag to stop run after confirm_changes
-        self.suppressed_summary: str = ""  # Store LLM summary to show after confirmation
-
-    async def from_agent_run_update(self, update: AgentResponseUpdate) -> list[BaseEvent]:
-        """
-        Convert an AgentResponseUpdate to AG-UI events.
-
-        Args:
-            update: The agent run update to convert.
-
-        Returns:
-            List of AG-UI events.
-        """
-        events: list[BaseEvent] = []
-
-        logger.info(f"Processing AgentRunUpdate with {len(update.contents)} content items")
-        for idx, content in enumerate(update.contents):
-            logger.info(f"  Content {idx}: type={type(content).__name__}")
-            if isinstance(content, TextContent):
-                events.extend(self._handle_text_content(content))
-            elif isinstance(content, FunctionCallContent):
-                events.extend(self._handle_function_call_content(content))
-            elif isinstance(content, FunctionResultContent):
-                events.extend(self._handle_function_result_content(content))
-            elif isinstance(content, FunctionApprovalRequestContent):
-                events.extend(self._handle_function_approval_request_content(content))
-
-        return events
-
-    def _handle_text_content(self, content: TextContent) -> list[BaseEvent]:
-        events: list[BaseEvent] = []
-        logger.info(f"  TextContent found: length={len(content.text)}")
-        logger.info(
-            "  Flags: skip_text_content=%s, should_stop_after_confirm=%s",
-            self.skip_text_content,
-            self.should_stop_after_confirm,
-        )
-
-        if self.skip_text_content:
-            logger.info("  SKIPPING TextContent: skip_text_content is True")
-            return events
-
-        if self.should_stop_after_confirm:
-            logger.info("  SKIPPING TextContent: waiting for confirm_changes response")
-            self.suppressed_summary += content.text
-            logger.info(f"  Suppressed summary length={len(self.suppressed_summary)}")
-            return events
-
-        # Skip empty text chunks to avoid emitting
-        # TextMessageContentEvent with an empty `delta` which fails
-        # Pydantic validation (AG-UI requires non-empty strings).
-        if not content.text:
-            logger.info("  SKIPPING TextContent: empty chunk")
-            return events
-
-        if not self.current_message_id:
-            self.current_message_id = generate_event_id()
-            start_event = TextMessageStartEvent(
-                message_id=self.current_message_id,
-                role="assistant",
-            )
-            logger.info(f"  EMITTING TextMessageStartEvent with message_id={self.current_message_id}")
-            events.append(start_event)
-
-        event = TextMessageContentEvent(
-            message_id=self.current_message_id,
-            delta=content.text,
-        )
-        logger.info(f"  EMITTING TextMessageContentEvent with text_len={len(content.text)}")
-        events.append(event)
-        return events
-
-    def _handle_function_call_content(self, content: FunctionCallContent) -> list[BaseEvent]:
-        events: list[BaseEvent] = []
-        if content.name:
-            logger.debug(f"Tool call: {content.name} (call_id: {content.call_id})")
-
-        if not content.name and not content.call_id and not self.current_tool_call_name:
-            args_length = len(str(content.arguments)) if content.arguments else 0
-            logger.warning(f"FunctionCallContent missing name and call_id. args_length={args_length}")
-
-        tool_call_id = self._coalesce_tool_call_id(content)
-        # Only emit ToolCallStartEvent once per tool call (when it's a new tool call)
-        if content.name and tool_call_id != self.current_tool_call_id:
-            self.streaming_tool_args = ""
-            self.state_delta_count = 0
-            self.current_tool_call_id = tool_call_id
-            self.current_tool_call_name = content.name
-
-            tool_start_event = ToolCallStartEvent(
-                tool_call_id=tool_call_id,
-                tool_call_name=content.name,
-                parent_message_id=self.current_message_id,
-            )
-            logger.info(f"Emitting ToolCallStartEvent with name='{content.name}', id='{tool_call_id}'")
-            events.append(tool_start_event)
-        elif tool_call_id:
-            self.current_tool_call_id = tool_call_id
-
-        if content.arguments:
-            delta_str = content.arguments if isinstance(content.arguments, str) else json.dumps(content.arguments)
-            logger.info(f"Emitting ToolCallArgsEvent with delta_length={len(delta_str)}, id='{tool_call_id}'")
-            args_event = ToolCallArgsEvent(
-                tool_call_id=tool_call_id,
-                delta=delta_str,
-            )
-            events.append(args_event)
-
-            events.extend(self._emit_predictive_state_deltas(delta_str))
-
-        return events
-
-    def _coalesce_tool_call_id(self, content: FunctionCallContent) -> str:
-        if content.call_id:
-            return content.call_id
-        if self.current_tool_call_id:
-            return self.current_tool_call_id
-        return generate_event_id()
-
-    def _emit_predictive_state_deltas(self, argument_chunk: str) -> list[BaseEvent]:
-        events: list[BaseEvent] = []
-        if not self.current_tool_call_name or not self.predict_state_config:
-            return events
-
-        self.streaming_tool_args += argument_chunk
-        logger.debug(
-            "Predictive state: accumulated %s chars for tool '%s'",
-            len(self.streaming_tool_args),
-            self.current_tool_call_name,
-        )
-
-        parsed_args = safe_json_parse(self.streaming_tool_args)
-        if parsed_args is None:
-            for state_key, config in self.predict_state_config.items():
-                if config["tool"] != self.current_tool_call_name:
-                    continue
-                tool_arg_name = config["tool_argument"]
-                pattern = rf'"{re.escape(tool_arg_name)}":\s*"([^"]*)'
-                match = re.search(pattern, self.streaming_tool_args)
-
-                if match:
-                    partial_value = match.group(1).replace("\\n", "\n").replace('\\"', '"').replace("\\\\", "\\")
-
-                    if state_key not in self.last_emitted_state or self.last_emitted_state[state_key] != partial_value:
-                        state_delta_event = StateDeltaEvent(
-                            delta=[
-                                {
-                                    "op": "replace",
-                                    "path": f"/{state_key}",
-                                    "value": partial_value,
-                                }
-                            ],
-                        )
-
-                        self.state_delta_count += 1
-                        if self.state_delta_count % 10 == 1:
-                            logger.info(
-                                "StateDeltaEvent #%s for '%s': op=replace, path=/%s, value_length=%s",
-                                self.state_delta_count,
-                                state_key,
-                                state_key,
-                                len(str(partial_value)),
-                            )
-                        elif self.state_delta_count % 100 == 0:
-                            logger.info(f"StateDeltaEvent #{self.state_delta_count} emitted")
-
-                        events.append(state_delta_event)
-                        self.last_emitted_state[state_key] = partial_value
-                        self.pending_state_updates[state_key] = partial_value
-
-        if parsed_args:
-            for state_key, config in self.predict_state_config.items():
-                if config["tool"] != self.current_tool_call_name:
-                    continue
-                tool_arg_name = config["tool_argument"]
-
-                state_value = extract_state_from_tool_args(parsed_args, tool_arg_name)
-                if state_value is None:
-                    continue
-
-                if state_key not in self.last_emitted_state or self.last_emitted_state[state_key] != state_value:
-                    state_delta_event = StateDeltaEvent(
-                        delta=[
-                            {
-                                "op": "replace",
-                                "path": f"/{state_key}",
-                                "value": state_value,
-                            }
-                        ],
-                    )
-
-                    self.state_delta_count += 1
-                    if self.state_delta_count % 10 == 1:
-                        logger.info(
-                            "StateDeltaEvent #%s for '%s': op=replace, path=/%s, value_length=%s",
-                            self.state_delta_count,
-                            state_key,
-                            state_key,
-                            len(str(state_value)),
-                        )
-                    elif self.state_delta_count % 100 == 0:
-                        logger.info(f"StateDeltaEvent #{self.state_delta_count} emitted")
-
-                    events.append(state_delta_event)
-                    self.last_emitted_state[state_key] = state_value
-                    self.pending_state_updates[state_key] = state_value
-        return events
-
-    def _handle_function_result_content(self, content: FunctionResultContent) -> list[BaseEvent]:
-        events: list[BaseEvent] = []
-        if content.call_id:
-            end_event = ToolCallEndEvent(
-                tool_call_id=content.call_id,
-            )
-            logger.info(f"Emitting ToolCallEndEvent for completed tool call '{content.call_id}'")
-            events.append(end_event)
-
-            if self.state_delta_count > 0:
-                logger.info(
-                    "Tool call '%s' complete: emitted %s StateDeltaEvents total",
-                    content.call_id,
-                    self.state_delta_count,
-                )
-
-            self.streaming_tool_args = ""
-            self.state_delta_count = 0
-
-        result_message_id = generate_event_id()
-        result_content = prepare_function_call_results(content.result)
-
-        result_event = ToolCallResultEvent(
-            message_id=result_message_id,
-            tool_call_id=content.call_id,
-            content=result_content,
-            role="tool",
-        )
-        events.append(result_event)
-        events.extend(self._emit_state_snapshot_and_confirmation())
-
-        return events
-
-    def _emit_state_snapshot_and_confirmation(self) -> list[BaseEvent]:
-        events: list[BaseEvent] = []
-        if self.pending_state_updates:
-            for key, value in self.pending_state_updates.items():
-                self.current_state[key] = value
-
-            logger.info(f"Emitting StateSnapshotEvent with keys: {list(self.current_state.keys())}")
-            if "recipe" in self.current_state:
-                recipe = self.current_state["recipe"]
-                logger.info(
-                    "Recipe fields: title=%s, skill_level=%s, ingredients_count=%s, instructions_count=%s",
-                    recipe.get("title"),
-                    recipe.get("skill_level"),
-                    len(recipe.get("ingredients", [])),
-                    len(recipe.get("instructions", [])),
-                )
-
-            state_snapshot_event = StateSnapshotEvent(
-                snapshot=self.current_state,
-            )
-            events.append(state_snapshot_event)
-
-            tool_was_predictive = False
-            logger.debug(
-                "Checking predictive state: current_tool='%s', predict_config=%s",
-                self.current_tool_call_name,
-                list(self.predict_state_config.keys()) if self.predict_state_config else "None",
-            )
-            for state_key, config in self.predict_state_config.items():
-                if self.current_tool_call_name and config["tool"] == self.current_tool_call_name:
-                    logger.info(
-                        "Tool '%s' matches predictive config for state key '%s'",
-                        self.current_tool_call_name,
-                        state_key,
-                    )
-                    tool_was_predictive = True
-                    break
-
-            if tool_was_predictive and self.require_confirmation:
-                events.extend(self._emit_confirm_changes_tool_call())
-            elif tool_was_predictive:
-                logger.info("Skipping confirm_changes - require_confirmation is False")
-
-            self.pending_state_updates.clear()
-            self.last_emitted_state = deepcopy(self.current_state)
-            self.current_tool_call_name = None
-        return events
-
-    def _emit_confirm_changes_tool_call(self, function_call: FunctionCallContent | None = None) -> list[BaseEvent]:
-        """Emit a confirm_changes tool call for Dojo UI compatibility.
-
-        Args:
-            function_call: Optional function call that needs confirmation.
-                If provided, includes function info in the confirm_changes args
-                so Dojo UI can display what's being confirmed.
-        """
-        events: list[BaseEvent] = []
-        confirm_call_id = generate_event_id()
-        logger.info("Emitting confirm_changes tool call for predictive update")
-
-        confirm_start = ToolCallStartEvent(
-            tool_call_id=confirm_call_id,
-            tool_call_name="confirm_changes",
-            parent_message_id=self.current_message_id,
-        )
-        events.append(confirm_start)
-
-        # Include function info if this is for a function approval
-        # This helps Dojo UI display meaningful confirmation info
-        if function_call:
-            args_dict = {
-                "function_name": function_call.name,
-                "function_call_id": function_call.call_id,
-                "function_arguments": function_call.parse_arguments() or {},
-                "steps": [
-                    {
-                        "description": f"Execute {function_call.name}",
-                        "status": "enabled",
-                    }
-                ],
-            }
-            args_json = json.dumps(args_dict)
-        else:
-            args_json = "{}"
-
-        confirm_args = ToolCallArgsEvent(
-            tool_call_id=confirm_call_id,
-            delta=args_json,
-        )
-        events.append(confirm_args)
-
-        confirm_end = ToolCallEndEvent(
-            tool_call_id=confirm_call_id,
-        )
-        events.append(confirm_end)
-
-        self.should_stop_after_confirm = True
-        logger.info("Set flag to stop run after confirm_changes")
-        return events
-
-    def _emit_function_approval_tool_call(self, function_call: FunctionCallContent) -> list[BaseEvent]:
-        """Emit a tool call that can drive UI approval for function requests."""
-        tool_call_name = "confirm_changes"
-        if self.approval_tool_name and self.approval_tool_name != function_call.name:
-            tool_call_name = self.approval_tool_name
-
-        tool_call_id = generate_event_id()
-        tool_start = ToolCallStartEvent(
-            tool_call_id=tool_call_id,
-            tool_call_name=tool_call_name,
-            parent_message_id=self.current_message_id,
-        )
-        events: list[BaseEvent] = [tool_start]
-
-        args_dict = {
-            "function_name": function_call.name,
-            "function_call_id": function_call.call_id,
-            "function_arguments": function_call.parse_arguments() or {},
-            "steps": [
-                {
-                    "description": f"Execute {function_call.name}",
-                    "status": "enabled",
-                }
-            ],
-        }
-        args_json = json.dumps(args_dict)
-
-        events.append(
-            ToolCallArgsEvent(
-                tool_call_id=tool_call_id,
-                delta=args_json,
-            )
-        )
-        events.append(
-            ToolCallEndEvent(
-                tool_call_id=tool_call_id,
-            )
-        )
-
-        self.should_stop_after_confirm = True
-        logger.info("Set flag to stop run after confirm_changes")
-        return events
-
-    def _handle_function_approval_request_content(self, content: FunctionApprovalRequestContent) -> list[BaseEvent]:
-        events: list[BaseEvent] = []
-        logger.info("=== FUNCTION APPROVAL REQUEST ===")
-        logger.info(f"  Function: {content.function_call.name}")
-        logger.info(f"  Call ID: {content.function_call.call_id}")
-
-        parsed_args = content.function_call.parse_arguments()
-        parsed_arg_keys = list(parsed_args.keys()) if parsed_args else "None"
-        logger.info(f"  Parsed args keys: {parsed_arg_keys}")
-
-        if parsed_args and self.predict_state_config:
-            logger.info(
-                "  Checking predict_state_config keys: %s",
-                list(self.predict_state_config.keys()) if self.predict_state_config else "None",
-            )
-            for state_key, config in self.predict_state_config.items():
-                if config["tool"] != content.function_call.name:
-                    continue
-                tool_arg_name = config["tool_argument"]
-                logger.info(
-                    "  MATCHED tool '%s' for state key '%s', arg='%s'",
-                    content.function_call.name,
-                    state_key,
-                    tool_arg_name,
-                )
-
-                state_value = extract_state_from_tool_args(parsed_args, tool_arg_name)
-                if state_value is None:
-                    logger.warning(f"  Tool argument '{tool_arg_name}' not found in parsed args")
-                    continue
-
-                self.current_state[state_key] = state_value
-                logger.info("Emitting StateSnapshotEvent for key '%s', value type: %s", state_key, type(state_value))  # type: ignore
-                state_snapshot = StateSnapshotEvent(
-                    snapshot=self.current_state,
-                )
-                events.append(state_snapshot)
-
-        if content.function_call.call_id:
-            end_event = ToolCallEndEvent(
-                tool_call_id=content.function_call.call_id,
-            )
-            logger.info(f"Emitting ToolCallEndEvent for approval-required tool '{content.function_call.call_id}'")
-            events.append(end_event)
-
-        # Emit the function_approval_request custom event for UI implementations that support it
-        approval_event = CustomEvent(
-            name="function_approval_request",
-            value={
-                "id": content.id,
-                "function_call": {
-                    "call_id": content.function_call.call_id,
-                    "name": content.function_call.name,
-                    "arguments": content.function_call.parse_arguments(),
-                },
-            },
-        )
-        logger.info(f"Emitting function_approval_request custom event for '{content.function_call.name}'")
-        events.append(approval_event)
-
-        # Emit a UI-friendly approval tool call for function approvals.
-        if self.require_confirmation:
-            events.extend(self._emit_function_approval_tool_call(content.function_call))
-
-        # Signal orchestrator to stop the run and wait for user approval response
-        self.should_stop_after_confirm = True
-        logger.info("Set flag to stop run - waiting for function approval response")
-        return events
-
-    def create_run_started_event(self) -> RunStartedEvent:
-        """Create a run started event."""
-        return RunStartedEvent(
-            run_id=self.run_id,
-            thread_id=self.thread_id,
-        )
-
-    def create_run_finished_event(self, result: Any = None) -> RunFinishedEvent:
-        """Create a run finished event."""
-        return RunFinishedEvent(
-            run_id=self.run_id,
-            thread_id=self.thread_id,
-            result=result,
-        )
-
-    def create_message_start_event(self, message_id: str, role: str = "assistant") -> TextMessageStartEvent:
-        """Create a message start event."""
-        return TextMessageStartEvent(
-            message_id=message_id,
-            role=role,  # type: ignore
-        )
-
-    def create_message_end_event(self, message_id: str) -> TextMessageEndEvent:
-        """Create a message end event."""
-        return TextMessageEndEvent(
-            message_id=message_id,
-        )
-
-    def create_state_snapshot_event(self, state: dict[str, Any]) -> StateSnapshotEvent:
-        """Create a state snapshot event.
-
-        Args:
-            state: The complete state snapshot.
-
-        Returns:
-            StateSnapshotEvent.
-        """
-        return StateSnapshotEvent(
-            snapshot=state,
-        )
-
-    def create_state_delta_event(self, delta: list[dict[str, Any]]) -> StateDeltaEvent:
-        """Create a state delta event using JSON Patch format (RFC 6902).
-
-        Args:
-            delta: List of JSON Patch operations.
-
-        Returns:
-            StateDeltaEvent.
-        """
-        return StateDeltaEvent(
-            delta=delta,
-        )
@@ -8,11 +8,8 @@ from typing import Any, cast

 from agent_framework import (
    ChatMessage,
-    FunctionApprovalResponseContent,
-    FunctionCallContent,
-    FunctionResultContent,
+    Content,
    Role,
-    TextContent,
    prepare_function_call_results,
 )

@@ -40,11 +37,11 @@ def _sanitize_tool_history(messages: list[ChatMessage]) -> list[ChatMessage]:
            tool_ids = {
                str(content.call_id)
                for content in msg.contents or []
-                if isinstance(content, FunctionCallContent) and content.call_id
+                if content.type == "function_call" and content.call_id
            }
            confirm_changes_call = None
            for content in msg.contents or []:
-                if isinstance(content, FunctionCallContent) and content.name == "confirm_changes":
+                if content.type == "function_call" and content.name == "confirm_changes":
                    confirm_changes_call = content
                    break

@@ -59,7 +56,7 @@ def _sanitize_tool_history(messages: list[ChatMessage]) -> list[ChatMessage]:
            approval_call_ids: set[str] = set()
            approval_accepted: bool | None = None
            for content in msg.contents or []:
-                if type(content) is FunctionApprovalResponseContent:
+                if content.type == "function_approval_response":
                    if content.function_call and content.function_call.call_id:
                        approval_call_ids.add(str(content.function_call.call_id))
                    if approval_accepted is None:
@@ -79,7 +76,7 @@ def _sanitize_tool_history(messages: list[ChatMessage]) -> list[ChatMessage]:
                synthetic_result = ChatMessage(
                    role="tool",
                    contents=[
-                        FunctionResultContent(
+                        Content.from_function_result(
                            call_id=pending_confirm_changes_id,
                            result="Confirmed" if approval_accepted else "Rejected",
                        )
@@ -93,12 +90,12 @@ def _sanitize_tool_history(messages: list[ChatMessage]) -> list[ChatMessage]:
            if pending_confirm_changes_id:
                user_text = ""
                for content in msg.contents or []:
-                    if isinstance(content, TextContent):
-                        user_text = content.text
+                    if content.type == "text":
+                        user_text = content.text  # type: ignore[assignment]
                        break

                try:
-                    parsed = json.loads(user_text)
+                    parsed = json.loads(user_text)  # type: ignore[arg-type]
                    if "accepted" in parsed:
                        logger.info(
                            f"Injecting synthetic tool result for confirm_changes call_id={pending_confirm_changes_id}"
@@ -106,7 +103,7 @@ def _sanitize_tool_history(messages: list[ChatMessage]) -> list[ChatMessage]:
                        synthetic_result = ChatMessage(
                            role="tool",
                            contents=[
-                                FunctionResultContent(
+                                Content.from_function_result(
                                    call_id=pending_confirm_changes_id,
                                    result="Confirmed" if parsed.get("accepted") else "Rejected",
                                )
@@ -130,7 +127,7 @@ def _sanitize_tool_history(messages: list[ChatMessage]) -> list[ChatMessage]:
                    synthetic_result = ChatMessage(
                        role="tool",
                        contents=[
-                            FunctionResultContent(
+                            Content.from_function_result(
                                call_id=pending_call_id,
                                result="Tool execution skipped - user provided follow-up message",
                            )
@@ -149,7 +146,7 @@ def _sanitize_tool_history(messages: list[ChatMessage]) -> list[ChatMessage]:
                continue
            keep = False
            for content in msg.contents or []:
-                if isinstance(content, FunctionResultContent):
+                if content.type == "function_result" and content.call_id:
                    call_id = str(content.call_id)
                    if call_id in pending_tool_call_ids:
                        keep = True
@@ -175,7 +172,7 @@ def _deduplicate_messages(messages: list[ChatMessage]) -> list[ChatMessage]:
    for idx, msg in enumerate(messages):
        role_value = get_role_value(msg)

-        if role_value == "tool" and msg.contents and isinstance(msg.contents[0], FunctionResultContent):
+        if role_value == "tool" and msg.contents and msg.contents[0].type == "function_result":
            call_id = str(msg.contents[0].call_id)
            key: Any = (role_value, call_id)

@@ -184,7 +181,7 @@ def _deduplicate_messages(messages: list[ChatMessage]) -> list[ChatMessage]:
                existing_msg = unique_messages[existing_idx]

                existing_result = None
-                if existing_msg.contents and isinstance(existing_msg.contents[0], FunctionResultContent):
+                if existing_msg.contents and existing_msg.contents[0].type == "function_result":
                    existing_result = existing_msg.contents[0].result
                new_result = msg.contents[0].result

@@ -198,11 +195,9 @@ def _deduplicate_messages(messages: list[ChatMessage]) -> list[ChatMessage]:
            seen_keys[key] = len(unique_messages)
            unique_messages.append(msg)

-        elif (
-            role_value == "assistant" and msg.contents and any(isinstance(c, FunctionCallContent) for c in msg.contents)
-        ):
+        elif role_value == "assistant" and msg.contents and any(c.type == "function_call" for c in msg.contents):
            tool_call_ids = tuple(
-                sorted(str(c.call_id) for c in msg.contents if isinstance(c, FunctionCallContent) and c.call_id)
+                sorted(str(c.call_id) for c in msg.contents if c.type == "function_call" and c.call_id)
            )
            key = (role_value, tool_call_ids)

@@ -257,33 +252,29 @@ def agui_messages_to_agent_framework(messages: list[dict[str, Any]]) -> list[Cha
            tool_calls = raw_msg.get("tool_calls") or raw_msg.get("toolCalls")
            if not isinstance(tool_calls, list):
                continue
-            tool_calls_list = cast(list[Any], tool_calls)
-            for tool_call in tool_calls_list:
+            for tool_call in tool_calls:
                if not isinstance(tool_call, dict):
                    continue
-                tool_call_dict = cast(dict[str, Any], tool_call)
-                if str(tool_call_dict.get("id", "")) != tool_call_id:
+                if str(tool_call.get("id", "")) != tool_call_id:
                    continue
-                function_payload = tool_call_dict.get("function")
+                function_payload = tool_call.get("function")
                if not isinstance(function_payload, dict):
                    return
-                function_payload_dict = cast(dict[str, Any], function_payload)
-                existing_args = function_payload_dict.get("arguments")
+                existing_args = function_payload.get("arguments")
                if isinstance(existing_args, str):
-                    function_payload_dict["arguments"] = json.dumps(modified_args)
+                    function_payload["arguments"] = json.dumps(modified_args)
                else:
-                    function_payload_dict["arguments"] = modified_args
+                    function_payload["arguments"] = modified_args
                return

-    def _find_matching_func_call(call_id: str) -> FunctionCallContent | None:
+    def _find_matching_func_call(call_id: str) -> Content | None:
        for prev_msg in result:
            role_val = prev_msg.role.value if hasattr(prev_msg.role, "value") else str(prev_msg.role)
            if role_val != "assistant":
                continue
            for content in prev_msg.contents or []:
-                if isinstance(content, FunctionCallContent):
-                    if content.call_id == call_id and content.name != "confirm_changes":
-                        return content
+                if content.type == "function_call" and content.call_id == call_id and content.name != "confirm_changes":
+                    return content
        return None

    def _parse_arguments(arguments: Any) -> dict[str, Any] | None:
@@ -301,9 +292,9 @@ def agui_messages_to_agent_framework(messages: list[dict[str, Any]]) -> list[Cha
                continue
            direct_call = None
            confirm_call = None
-            sibling_calls: list[FunctionCallContent] = []
+            sibling_calls: list[Content] = []
            for content in prev_msg.contents or []:
-                if not isinstance(content, FunctionCallContent):
+                if content.type != "function_call":
                    continue
                if content.call_id == tool_call_id:
                    direct_call = content
@@ -407,7 +398,7 @@ def agui_messages_to_agent_framework(messages: list[dict[str, Any]]) -> list[Cha
                        if not (
                            (m.role.value if hasattr(m.role, "value") else str(m.role)) == "tool"
                            and any(
-                                isinstance(c, FunctionResultContent) and c.call_id == approval_call_id
+                                c.type == "function_result" and c.call_id == approval_call_id
                                for c in (m.contents or [])
                            )
                        )
@@ -439,8 +430,7 @@ def agui_messages_to_agent_framework(messages: list[dict[str, Any]]) -> list[Cha
                                        if desc:
                                            approved_by_description[str(desc)] = step_item_dict
                                merged_steps: list[Any] = []
-                                original_steps_list = cast(list[Any], original_steps)
-                                for orig_step in original_steps_list:
+                                for orig_step in original_steps:
                                    if not isinstance(orig_step, dict):
                                        merged_steps.append(orig_step)
                                        continue
@@ -465,9 +455,9 @@ def agui_messages_to_agent_framework(messages: list[dict[str, Any]]) -> list[Cha
                        matching_func_call.arguments = updated_args
                        _update_tool_call_arguments(messages, str(approval_call_id), merged_args)
                        # Create a new FunctionCallContent with the modified arguments
-                        func_call_for_approval = FunctionCallContent(
-                            call_id=matching_func_call.call_id,
-                            name=matching_func_call.name,
+                        func_call_for_approval = Content.from_function_call(
+                            call_id=matching_func_call.call_id,  # type: ignore[arg-type]
+                            name=matching_func_call.name,  # type: ignore[arg-type]
                            arguments=json.dumps(filtered_args),
                        )
                        logger.info(f"Using modified arguments from approval: {filtered_args}")
@@ -476,7 +466,7 @@ def agui_messages_to_agent_framework(messages: list[dict[str, Any]]) -> list[Cha
                        func_call_for_approval = matching_func_call

                    # Create FunctionApprovalResponseContent for the agent framework
-                    approval_response = FunctionApprovalResponseContent(
+                    approval_response = Content.from_function_approval_response(
                        approved=accepted,
                        id=str(approval_call_id),
                        function_call=func_call_for_approval,
@@ -491,7 +481,7 @@ def agui_messages_to_agent_framework(messages: list[dict[str, Any]]) -> list[Cha
                    # Keep the old behavior for backwards compatibility
                    chat_msg = ChatMessage(
                        role=Role.USER,
-                        contents=[TextContent(text=approval_payload_text)],
+                        contents=[Content.from_text(text=approval_payload_text)],
                        additional_properties={"is_tool_result": True, "tool_call_id": str(tool_call_id or "")},
                    )
                if "id" in msg:
@@ -504,14 +494,14 @@ def agui_messages_to_agent_framework(messages: list[dict[str, Any]]) -> list[Cha
            if isinstance(result_content, str):
                func_result = result_content
            elif isinstance(result_content, dict):
-                func_result = cast(dict[str, Any], result_content)
+                func_result = result_content
            elif isinstance(result_content, list):
-                func_result = cast(list[Any], result_content)
+                func_result = result_content
            else:
                func_result = str(result_content)
            chat_msg = ChatMessage(
                role=Role.TOOL,
-                contents=[FunctionResultContent(call_id=str(tool_call_id), result=func_result)],
+                contents=[Content.from_function_result(call_id=str(tool_call_id), result=func_result)],
            )
            if "id" in msg:
                chat_msg.message_id = msg["id"]
@@ -527,21 +517,21 @@ def agui_messages_to_agent_framework(messages: list[dict[str, Any]]) -> list[Cha

            chat_msg = ChatMessage(
                role=Role.TOOL,
-                contents=[FunctionResultContent(call_id=str(tool_call_id), result=result_content)],
+                contents=[Content.from_function_result(call_id=str(tool_call_id), result=result_content)],
            )
            if "id" in msg:
                chat_msg.message_id = msg["id"]
            result.append(chat_msg)
            continue

-        # If assistant message includes tool calls, convert to FunctionCallContent(s)
+        # If assistant message includes tool calls, convert to Content.from_function_call(s)
        tool_calls = msg.get("tool_calls") or msg.get("toolCalls")
        if tool_calls:
            contents: list[Any] = []
            # Include any assistant text content if present
            content_text = msg.get("content")
            if isinstance(content_text, str) and content_text:
-                contents.append(TextContent(text=content_text))
+                contents.append(Content.from_text(text=content_text))
            # Convert each tool call entry
            for tc in tool_calls:
                if not isinstance(tc, dict):
@@ -558,7 +548,7 @@ def agui_messages_to_agent_framework(messages: list[dict[str, Any]]) -> list[Cha
                    arguments = func_dict.get("arguments")

                    contents.append(
-                        FunctionCallContent(
+                        Content.from_function_call(
                            call_id=call_id,
                            name=name,
                            arguments=arguments,
@@ -580,14 +570,14 @@ def agui_messages_to_agent_framework(messages: list[dict[str, Any]]) -> list[Cha
            approval_contents: list[Any] = []
            for approval in msg["function_approvals"]:
                # Create FunctionCallContent with the modified arguments
-                func_call = FunctionCallContent(
+                func_call = Content.from_function_call(
                    call_id=approval.get("call_id", ""),
                    name=approval.get("name", ""),
                    arguments=approval.get("arguments", {}),
                )

                # Create the approval response
-                approval_response = FunctionApprovalResponseContent(
+                approval_response = Content.from_function_approval_response(
                    approved=approval.get("approved", True),
                    id=approval.get("id", ""),
                    function_call=func_call,
@@ -599,9 +589,9 @@ def agui_messages_to_agent_framework(messages: list[dict[str, Any]]) -> list[Cha
            # Regular text message
            content = msg.get("content", "")
            if isinstance(content, str):
-                chat_msg = ChatMessage(role=role, contents=[TextContent(text=content)])
+                chat_msg = ChatMessage(role=role, contents=[Content.from_text(text=content)])
            else:
-                chat_msg = ChatMessage(role=role, contents=[TextContent(text=str(content))])
+                chat_msg = ChatMessage(role=role, contents=[Content.from_text(text=str(content))])

        if "id" in msg:
            chat_msg.message_id = msg["id"]
@@ -652,9 +642,9 @@ def agent_framework_messages_to_agui(messages: list[ChatMessage] | list[dict[str
        tool_result_call_id: str | None = None

        for content in msg.contents:
-            if isinstance(content, TextContent):
-                content_text += content.text
-            elif isinstance(content, FunctionCallContent):
+            if content.type == "text":
+                content_text += content.text  # type: ignore[operator]
+            elif content.type == "function_call":
                tool_calls.append(
                    {
                        "id": content.call_id,
@@ -665,7 +655,7 @@ def agent_framework_messages_to_agui(messages: list[ChatMessage] | list[dict[str
                        },
                    }
                )
-            elif isinstance(content, FunctionResultContent):
+            elif content.type == "function_result":
                # Tool result content - extract call_id and result
                tool_result_call_id = content.call_id
                # Serialize result to string using core utility
@@ -702,8 +692,13 @@ def extract_text_from_contents(contents: list[Any]) -> str:
    """
    text_parts: list[str] = []
    for content in contents:
-        if isinstance(content, TextContent):
-            text_parts.append(content.text)
+        if type_ := getattr(content, "type", None):
+            if type_ == "text_reasoning":
+                continue
+            if text := getattr(content, "text", None):
+                text_parts.append(text)
+            continue
+        # TODO (moonbox3): should this handle both text and text_reasoning?
        elif hasattr(content, "text"):
            text_parts.append(content.text)
    return "".join(text_parts)
@@ -735,40 +730,35 @@ def agui_messages_to_snapshot_format(messages: list[dict[str, Any]]) -> list[dic
        if isinstance(content, list):
            # Convert content array format to simple string
            text_parts: list[str] = []
-            content_list = cast(list[Any], content)
-            for item in content_list:
+            for item in content:
                if isinstance(item, dict):
-                    item_dict = cast(dict[str, Any], item)
                    # Convert 'input_text' to 'text' type
-                    if item_dict.get("type") == "input_text":
-                        text_parts.append(str(item_dict.get("text", "")))
-                    elif item_dict.get("type") == "text":
-                        text_parts.append(str(item_dict.get("text", "")))
+                    if item.get("type") == "input_text":
+                        text_parts.append(str(item.get("text", "")))
+                    elif item.get("type") == "text":
+                        text_parts.append(str(item.get("text", "")))
                    else:
                        # Other types - just extract text field if present
-                        text_parts.append(str(item_dict.get("text", "")))
+                        text_parts.append(str(item.get("text", "")))
            normalized_msg["content"] = "".join(text_parts)
        elif content is None:
            normalized_msg["content"] = ""

        tool_calls = normalized_msg.get("tool_calls") or normalized_msg.get("toolCalls")
        if isinstance(tool_calls, list):
-            tool_calls_list = cast(list[Any], tool_calls)
-            for tool_call in tool_calls_list:
+            for tool_call in tool_calls:
                if not isinstance(tool_call, dict):
                    continue
-                tool_call_dict = cast(dict[str, Any], tool_call)
-                function_payload = tool_call_dict.get("function")
+                function_payload = tool_call.get("function")
                if not isinstance(function_payload, dict):
                    continue
-                function_payload_dict = cast(dict[str, Any], function_payload)
-                if "arguments" not in function_payload_dict:
+                if "arguments" not in function_payload:
                    continue
-                arguments = function_payload_dict.get("arguments")
+                arguments = function_payload.get("arguments")
                if arguments is None:
-                    function_payload_dict["arguments"] = ""
+                    function_payload["arguments"] = ""
                elif not isinstance(arguments, str):
-                    function_payload_dict["arguments"] = json.dumps(arguments)
+                    function_payload["arguments"] = json.dumps(arguments)

        # Normalize tool_call_id to toolCallId for tool messages
        normalized_msg["role"] = normalize_agui_role(normalized_msg.get("role"))
@@ -782,11 +772,3 @@ def agui_messages_to_snapshot_format(messages: list[dict[str, Any]]) -> list[dic
        result.append(normalized_msg)

    return result
-
-
-__all__ = [
-    "agui_messages_to_agent_framework",
-    "agent_framework_messages_to_agui",
-    "agui_messages_to_snapshot_format",
-    "extract_text_from_contents",
-]
@@ -1,25 +1,21 @@
 # Copyright (c) Microsoft. All rights reserved.

-"""Helper functions for orchestration logic."""
+"""Helper functions for orchestration logic.
+
+Most orchestration helpers have been moved inline to _run.py.
+This module retains utilities that may be useful for testing or extensions.
+"""

 import json
 import logging
-from typing import TYPE_CHECKING, Any
+from typing import Any

-from ag_ui.core import StateSnapshotEvent
 from agent_framework import (
    ChatMessage,
-    FunctionApprovalResponseContent,
-    FunctionCallContent,
-    FunctionResultContent,
-    TextContent,
+    Content,
 )

-from .._utils import get_role_value, safe_json_parse
-
-if TYPE_CHECKING:
-    from .._events import AgentFrameworkEventBridge
-    from ._state_manager import StateManager
+from .._utils import get_role_value

 logger = logging.getLogger(__name__)

@@ -37,9 +33,9 @@ def pending_tool_call_ids(messages: list[ChatMessage]) -> set[str]:
    resolved_ids: set[str] = set()
    for msg in messages:
        for content in msg.contents:
-            if isinstance(content, FunctionCallContent) and content.call_id:
+            if content.type == "function_call" and content.call_id:
                pending_ids.add(str(content.call_id))
-            elif isinstance(content, FunctionResultContent) and content.call_id:
+            elif content.type == "function_result" and content.call_id:
                resolved_ids.add(str(content.call_id))
    return pending_ids - resolved_ids

@@ -56,7 +52,7 @@ def is_state_context_message(message: ChatMessage) -> bool:
    if get_role_value(message) != "system":
        return False
    for content in message.contents:
-        if isinstance(content, TextContent) and content.text.startswith("Current state of the application:"):
+        if content.type == "text" and content.text.startswith("Current state of the application:"):  # type: ignore[union-attr]
            return True
    return False

@@ -114,53 +110,6 @@ def tool_name_for_call_id(
    return str(name) if name else None


-def tool_calls_match_state(
-    provider_messages: list[ChatMessage],
-    state_manager: "StateManager",
-) -> bool:
-    """Check if tool calls in messages match current state.
-
-    Args:
-        provider_messages: Messages to check
-        state_manager: State manager with config and current state
-
-    Returns:
-        True if tool calls match state configuration
-    """
-    if not state_manager.predict_state_config or not state_manager.current_state:
-        return False
-
-    for state_key, config in state_manager.predict_state_config.items():
-        tool_name = config["tool"]
-        tool_arg_name = config["tool_argument"]
-        tool_args: dict[str, Any] | None = None
-
-        for msg in reversed(provider_messages):
-            if get_role_value(msg) != "assistant":
-                continue
-            for content in msg.contents:
-                if isinstance(content, FunctionCallContent) and content.name == tool_name:
-                    tool_args = safe_json_parse(content.arguments)
-                    break
-            if tool_args is not None:
-                break
-
-        if not tool_args:
-            return False
-
-        if tool_arg_name == "*":
-            state_value = tool_args
-        elif tool_arg_name in tool_args:
-            state_value = tool_args[tool_arg_name]
-        else:
-            return False
-
-        if state_manager.current_state.get(state_key) != state_value:
-            return False
-
-    return True
-
-
 def schema_has_steps(schema: Any) -> bool:
    """Check if a schema has a steps array property.

@@ -205,45 +154,10 @@ def select_approval_tool_name(client_tools: list[Any] | None) -> str | None:
    return None


-def select_messages_to_run(
-    provider_messages: list[ChatMessage],
-    state_manager: "StateManager",
-) -> list[ChatMessage]:
-    """Select and prepare messages for agent execution.
-
-    Injects state context message when appropriate.
-
-    Args:
-        provider_messages: Original messages from client
-        state_manager: State manager instance
-
-    Returns:
-        Messages ready for agent execution
-    """
-    if not provider_messages:
-        return []
-
-    is_new_user_turn = get_role_value(provider_messages[-1]) == "user"
-    conversation_has_tool_calls = tool_calls_match_state(provider_messages, state_manager)
-    state_context_msg = state_manager.state_context_message(
-        is_new_user_turn=is_new_user_turn, conversation_has_tool_calls=conversation_has_tool_calls
-    )
-    if not state_context_msg:
-        return list(provider_messages)
-
-    messages_to_run = [msg for msg in provider_messages if not is_state_context_message(msg)]
-    if pending_tool_call_ids(messages_to_run):
-        return messages_to_run
-
-    insert_index = len(messages_to_run) - 1 if is_new_user_turn else len(messages_to_run)
-    if insert_index < 0:
-        insert_index = 0
-    messages_to_run.insert(insert_index, state_context_msg)
-    return messages_to_run
-
-
 def build_safe_metadata(thread_metadata: dict[str, Any] | None) -> dict[str, Any]:
-    """Build metadata dict with truncated string values.
+    """Build metadata dict with truncated string values for Azure compatibility.
+
+    Azure has a 512 character limit per metadata value.

    Args:
        thread_metadata: Raw metadata dict
@@ -262,64 +176,7 @@ def build_safe_metadata(thread_metadata: dict[str, Any] | None) -> dict[str, Any
    return safe_metadata


-def collect_approved_state_snapshots(
-    provider_messages: list[ChatMessage],
-    predict_state_config: dict[str, dict[str, str]] | None,
-    current_state: dict[str, Any],
-    event_bridge: "AgentFrameworkEventBridge",
-) -> list[StateSnapshotEvent]:
-    """Collect state snapshots from approved function calls.
-
-    Args:
-        provider_messages: Messages containing approvals
-        predict_state_config: Predictive state configuration
-        current_state: Current state dict (will be mutated)
-        event_bridge: Event bridge for creating events
-
-    Returns:
-        List of state snapshot events
-    """
-    if not predict_state_config:
-        return []
-
-    events: list[StateSnapshotEvent] = []
-    for msg in provider_messages:
-        if get_role_value(msg) != "user":
-            continue
-        for content in msg.contents:
-            if type(content) is FunctionApprovalResponseContent:
-                if not content.function_call or not content.approved:
-                    continue
-                parsed_args = content.function_call.parse_arguments()
-                state_args = None
-                if content.additional_properties:
-                    state_args = content.additional_properties.get("ag_ui_state_args")
-                if not isinstance(state_args, dict):
-                    state_args = parsed_args
-                if not state_args:
-                    continue
-                for state_key, config in predict_state_config.items():
-                    if config["tool"] != content.function_call.name:
-                        continue
-                    tool_arg_name = config["tool_argument"]
-                    if tool_arg_name == "*":
-                        state_value = state_args
-                    elif isinstance(state_args, dict) and tool_arg_name in state_args:
-                        state_value = state_args[tool_arg_name]
-                    else:
-                        continue
-                    current_state[state_key] = state_value
-                    event_bridge.current_state[state_key] = state_value
-                    logger.info(
-                        f"Emitting StateSnapshotEvent for approved state key '{state_key}' "
-                        f"with {len(state_value) if isinstance(state_value, list) else 'N/A'} items"
-                    )
-                    events.append(StateSnapshotEvent(snapshot=current_state))
-                    break
-    return events
-
-
-def latest_approval_response(messages: list[ChatMessage]) -> FunctionApprovalResponseContent | None:
+def latest_approval_response(messages: list[ChatMessage]) -> Content | None:
    """Get the latest approval response from messages.

    Args:
@@ -332,12 +189,12 @@ def latest_approval_response(messages: list[ChatMessage]) -> FunctionApprovalRes
        return None
    last_message = messages[-1]
    for content in last_message.contents:
-        if type(content) is FunctionApprovalResponseContent:
+        if content.type == "function_approval_response":
            return content
    return None


-def approval_steps(approval: FunctionApprovalResponseContent) -> list[Any]:
+def approval_steps(approval: Content) -> list[Any]:
    """Extract steps from an approval response.

    Args:
@@ -346,9 +203,7 @@ def approval_steps(approval: FunctionApprovalResponseContent) -> list[Any]:
    Returns:
        List of steps, or empty list if none
    """
-    state_args: Any | None = None
-    if approval.additional_properties:
-        state_args = approval.additional_properties.get("ag_ui_state_args")
+    state_args = approval.additional_properties.get("ag_ui_state_args", None)
    if isinstance(state_args, dict):
        steps = state_args.get("steps")
        if isinstance(steps, list):
@@ -365,7 +220,7 @@ def approval_steps(approval: FunctionApprovalResponseContent) -> list[Any]:


 def is_step_based_approval(
-    approval: FunctionApprovalResponseContent,
+    approval: Content,
    predict_state_config: dict[str, dict[str, str]] | None,
 ) -> bool:
    """Check if an approval is step-based.
@@ -1,106 +0,0 @@
-# Copyright (c) Microsoft. All rights reserved.
-
-"""State orchestration utilities."""
-
-import json
-from typing import Any
-
-from ag_ui.core import CustomEvent, EventType
-from agent_framework import ChatMessage, TextContent
-
-
-class StateManager:
-    """Coordinates state defaults, snapshots, and structured updates."""
-
-    def __init__(
-        self,
-        state_schema: dict[str, Any] | None,
-        predict_state_config: dict[str, dict[str, str]] | None,
-        require_confirmation: bool,
-    ) -> None:
-        self.state_schema = state_schema or {}
-        self.predict_state_config = predict_state_config or {}
-        self.require_confirmation = require_confirmation
-        self.current_state: dict[str, Any] = {}
-        self._state_from_input: bool = False
-
-    def initialize(self, initial_state: dict[str, Any] | None) -> dict[str, Any]:
-        """Initialize state with schema defaults."""
-        self._state_from_input = initial_state is not None
-        self.current_state = (initial_state or {}).copy()
-        self._apply_schema_defaults()
-        return self.current_state
-
-    def predict_state_event(self) -> CustomEvent | None:
-        """Create predict-state custom event when configured."""
-        if not self.predict_state_config:
-            return None
-
-        predict_state_value = [
-            {
-                "state_key": state_key,
-                "tool": config["tool"],
-                "tool_argument": config["tool_argument"],
-            }
-            for state_key, config in self.predict_state_config.items()
-        ]
-
-        return CustomEvent(
-            type=EventType.CUSTOM,
-            name="PredictState",
-            value=predict_state_value,
-        )
-
-    def initial_snapshot_event(self, event_bridge: Any) -> Any:
-        """Emit initial snapshot when schema and state present."""
-        if not self.state_schema:
-            return None
-        self._apply_schema_defaults()
-        return event_bridge.create_state_snapshot_event(self.current_state)
-
-    def state_context_message(self, is_new_user_turn: bool, conversation_has_tool_calls: bool) -> ChatMessage | None:
-        """Inject state context only when starting a new user turn."""
-        if not self.current_state or not self.state_schema:
-            return None
-        if not is_new_user_turn:
-            return None
-        if conversation_has_tool_calls and not self._state_from_input:
-            return None
-
-        state_json = json.dumps(self.current_state, indent=2)
-        return ChatMessage(
-            role="system",
-            contents=[
-                TextContent(
-                    text=(
-                        "Current state of the application:\n"
-                        f"{state_json}\n\n"
-                        "When modifying state, you MUST include ALL existing data plus your changes.\n"
-                        "For example, if adding one new item to a list, include ALL existing items PLUS the one new item.\n"
-                        "Never replace existing data - always preserve and append or merge."
-                    )
-                )
-            ],
-        )
-
-    def extract_state_updates(self, response_dict: dict[str, Any]) -> dict[str, Any]:
-        """Extract state updates from structured response payloads."""
-        if self.state_schema:
-            return {key: response_dict[key] for key in self.state_schema.keys() if key in response_dict}
-        return {k: v for k, v in response_dict.items() if k != "message"}
-
-    def apply_state_updates(self, updates: dict[str, Any]) -> None:
-        """Merge state updates into current state."""
-        if not updates:
-            return
-        self.current_state.update(updates)
-
-    def _apply_schema_defaults(self) -> None:
-        """Fill missing state fields based on schema hints."""
-        for key, schema in self.state_schema.items():
-            if key in self.current_state:
-                continue
-            if isinstance(schema, dict) and schema.get("type") == "array":  # type: ignore
-                self.current_state[key] = []
-            else:
-                self.current_state[key] = {}
@@ -84,9 +84,26 @@ def register_additional_client_tools(agent: "AgentProtocol", client_tools: list[
        logger.debug(f"[TOOLS] Registered {len(client_tools)} client tools as additional_tools (declaration-only)")


+def _has_approval_tools(tools: list[Any]) -> bool:
+    """Check if any tools require approval."""
+    return any(getattr(tool, "approval_mode", None) == "always_require" for tool in tools)
+
+
 def merge_tools(server_tools: list[Any], client_tools: list[Any] | None) -> list[Any] | None:
-    """Combine server and client tools without overriding server metadata."""
+    """Combine server and client tools without overriding server metadata.
+
+    IMPORTANT: When server tools have approval_mode="always_require", we MUST return
+    them so they get passed to the streaming response handler. Otherwise, the approval
+    check in _try_execute_function_calls won't find the tool and won't trigger approval.
+    """
    if not client_tools:
+        # Even without client tools, we must pass server tools if any require approval
+        if server_tools and _has_approval_tools(server_tools):
+            logger.info(
+                f"[TOOLS] No client tools but server has approval tools - "
+                f"passing {len(server_tools)} server tools for approval mode"
+            )
+            return server_tools
        logger.info("[TOOLS] No client tools - not passing tools= parameter (using agent's configured tools)")
        return None

@@ -94,6 +111,13 @@ def merge_tools(server_tools: list[Any], client_tools: list[Any] | None) -> list
    unique_client_tools = [tool for tool in client_tools if getattr(tool, "name", None) not in server_tool_names]

    if not unique_client_tools:
+        # Same check: must pass server tools if any require approval
+        if server_tools and _has_approval_tools(server_tools):
+            logger.info(
+                f"[TOOLS] Client tools duplicate server but server has approval tools - "
+                f"passing {len(server_tools)} server tools for approval mode"
+            )
+            return server_tools
        logger.info("[TOOLS] All client tools duplicate server tools - not passing tools= parameter")
        return None

@@ -1,807 +0,0 @@
-# Copyright (c) Microsoft. All rights reserved.
-
-"""Orchestrators for multi-turn agent flows."""
-
-import json
-import logging
-import uuid
-from abc import ABC, abstractmethod
-from collections.abc import AsyncGenerator, Sequence
-from typing import TYPE_CHECKING, Any
-
-from ag_ui.core import (
-    BaseEvent,
-    MessagesSnapshotEvent,
-    RunErrorEvent,
-    TextMessageContentEvent,
-    TextMessageEndEvent,
-    TextMessageStartEvent,
-    ToolCallArgsEvent,
-    ToolCallEndEvent,
-    ToolCallResultEvent,
-    ToolCallStartEvent,
-)
-from agent_framework import (
-    AgentProtocol,
-    AgentThread,
-    ChatAgent,
-    FunctionCallContent,
-    FunctionResultContent,
-    TextContent,
-)
-from agent_framework._middleware import extract_and_merge_function_middleware
-from agent_framework._tools import (
-    FunctionInvocationConfiguration,
-    _collect_approval_responses,  # type: ignore
-    _replace_approval_contents_with_results,  # type: ignore
-    _try_execute_function_calls,  # type: ignore
-)
-
-from ._orchestration._helpers import (
-    approval_steps,
-    build_safe_metadata,
-    collect_approved_state_snapshots,
-    ensure_tool_call_entry,
-    is_step_based_approval,
-    latest_approval_response,
-    select_approval_tool_name,
-    select_messages_to_run,
-    tool_name_for_call_id,
-)
-from ._orchestration._tooling import (
-    collect_server_tools,
-    merge_tools,
-    register_additional_client_tools,
-)
-from ._utils import (
-    convert_agui_tools_to_agent_framework,
-    generate_event_id,
-    get_conversation_id_from_update,
-    get_role_value,
-)
-
-if TYPE_CHECKING:
-    from ._agent import AgentConfig
-    from ._confirmation_strategies import ConfirmationStrategy
-    from ._events import AgentFrameworkEventBridge
-    from ._orchestration._state_manager import StateManager
-
-
-logger = logging.getLogger(__name__)
-
-
-class ExecutionContext:
-    """Shared context for orchestrators."""
-
-    def __init__(
-        self,
-        input_data: dict[str, Any],
-        agent: AgentProtocol,
-        config: "AgentConfig",  # noqa: F821
-        confirmation_strategy: "ConfirmationStrategy | None" = None,  # noqa: F821
-    ):
-        """Initialize execution context.
-
-        Args:
-            input_data: AG-UI run input containing messages, state, etc.
-            agent: The Agent Framework agent to execute
-            config: Agent configuration
-            confirmation_strategy: Strategy for generating confirmation messages
-        """
-        self.input_data = input_data
-        self.agent = agent
-        self.config = config
-        self.confirmation_strategy = confirmation_strategy
-
-        # Lazy-loaded properties
-        self._messages = None
-        self._snapshot_messages = None
-        self._last_message = None
-        self._run_id: str | None = None
-        self._thread_id: str | None = None
-        self._supplied_run_id: str | None = None
-        self._supplied_thread_id: str | None = None
-
-    @property
-    def messages(self):
-        """Get converted Agent Framework messages (lazy loaded)."""
-        if self._messages is None:
-            from ._message_adapters import normalize_agui_input_messages
-
-            raw = self.input_data.get("messages", [])
-            if not isinstance(raw, list):
-                raw = []
-            self._messages, self._snapshot_messages = normalize_agui_input_messages(raw)
-        return self._messages
-
-    @property
-    def snapshot_messages(self) -> list[dict[str, Any]]:
-        """Get normalized AG-UI snapshot messages (lazy loaded)."""
-        if self._snapshot_messages is None:
-            if self._messages is None:
-                _ = self.messages
-            else:
-                from ._message_adapters import agent_framework_messages_to_agui, agui_messages_to_snapshot_format
-
-                raw_snapshot = agent_framework_messages_to_agui(self._messages)
-                self._snapshot_messages = agui_messages_to_snapshot_format(raw_snapshot)
-        return self._snapshot_messages or []
-
-    @property
-    def last_message(self):
-        """Get the last message in the conversation (lazy loaded)."""
-        if self._last_message is None and self.messages:
-            self._last_message = self.messages[-1]
-        return self._last_message
-
-    @property
-    def supplied_run_id(self) -> str | None:
-        """Get the supplied run ID, if any."""
-        if self._supplied_run_id is None:
-            self._supplied_run_id = self.input_data.get("run_id") or self.input_data.get("runId")
-        return self._supplied_run_id
-
-    @property
-    def run_id(self) -> str:
-        """Get supplied run ID or generate a new run ID."""
-        if self._run_id:
-            return self._run_id
-
-        if self.supplied_run_id:
-            self._run_id = self.supplied_run_id
-
-        if self._run_id is None:
-            self._run_id = str(uuid.uuid4())
-
-        return self._run_id
-
-    @property
-    def supplied_thread_id(self) -> str | None:
-        """Get the supplied thread ID, if any."""
-        if self._supplied_thread_id is None:
-            self._supplied_thread_id = self.input_data.get("thread_id") or self.input_data.get("threadId")
-        return self._supplied_thread_id
-
-    @property
-    def thread_id(self) -> str:
-        """Get supplied thread ID or generate a new thread ID."""
-        if self._thread_id:
-            return self._thread_id
-
-        if self.supplied_thread_id:
-            self._thread_id = self.supplied_thread_id
-
-        if self._thread_id is None:
-            self._thread_id = str(uuid.uuid4())
-
-        return self._thread_id
-
-    def update_run_id(self, new_run_id: str) -> None:
-        """Update the run ID in the context.
-
-        Args:
-            new_run_id: The new run ID to set
-        """
-        self._supplied_run_id = new_run_id
-        self._run_id = new_run_id
-
-    def update_thread_id(self, new_thread_id: str) -> None:
-        """Update the thread ID in the context.
-
-        Args:
-            new_thread_id: The new thread ID to set
-        """
-        self._supplied_thread_id = new_thread_id
-        self._thread_id = new_thread_id
-
-
-class Orchestrator(ABC):
-    """Base orchestrator for agent execution flows."""
-
-    @abstractmethod
-    def can_handle(self, context: ExecutionContext) -> bool:
-        """Determine if this orchestrator handles the current request.
-
-        Args:
-            context: Execution context with input data and agent
-
-        Returns:
-            True if this orchestrator should handle the request
-        """
-        ...
-
-    @abstractmethod
-    async def run(
-        self,
-        context: ExecutionContext,
-    ) -> AsyncGenerator[BaseEvent, None]:
-        """Execute the orchestration and yield events.
-
-        Args:
-            context: Execution context
-
-        Yields:
-            AG-UI events
-        """
-        # This is never executed - just satisfies mypy's requirement for async generators
-        if False:  # pragma: no cover
-            yield
-        raise NotImplementedError
-
-
-class HumanInTheLoopOrchestrator(Orchestrator):
-    """Handles tool approval responses from user."""
-
-    def can_handle(self, context: ExecutionContext) -> bool:
-        """Check if last message is a tool approval response.
-
-        Args:
-            context: Execution context
-
-        Returns:
-            True if last message is a tool result
-        """
-        msg = context.last_message
-        if not msg:
-            return False
-
-        return bool(msg.additional_properties.get("is_tool_result", False))
-
-    async def run(
-        self,
-        context: ExecutionContext,
-    ) -> AsyncGenerator[BaseEvent, None]:
-        """Process approval response and generate confirmation events.
-
-        This implementation is extracted from the legacy _agent.py lines 144-244.
-
-        Args:
-            context: Execution context
-
-        Yields:
-            AG-UI events (TextMessage, RunFinished)
-        """
-        from ._confirmation_strategies import DefaultConfirmationStrategy
-        from ._events import AgentFrameworkEventBridge
-
-        logger.info("=== TOOL RESULT DETECTED (HumanInTheLoopOrchestrator) ===")
-
-        # Create event bridge for run events
-        event_bridge = AgentFrameworkEventBridge(
-            run_id=context.run_id,
-            thread_id=context.thread_id,
-        )
-
-        # CRITICAL: Every AG-UI run must start with RunStartedEvent
-        yield event_bridge.create_run_started_event()
-
-        # Get confirmation strategy (use default if none provided)
-        strategy = context.confirmation_strategy
-        if strategy is None:
-            strategy = DefaultConfirmationStrategy()
-
-        # Parse the tool result content
-        tool_content_text = ""
-        last_message = context.last_message
-        if last_message:
-            for content in last_message.contents:
-                if isinstance(content, TextContent):
-                    tool_content_text = content.text
-                    break
-
-        try:
-            tool_result = json.loads(tool_content_text)
-            accepted = tool_result.get("accepted", False)
-            steps = tool_result.get("steps", [])
-
-            logger.info(f"  Accepted: {accepted}")
-            logger.info(f"  Steps count: {len(steps)}")
-
-            # Emit a text message confirming execution
-            message_id = generate_event_id()
-
-            yield TextMessageStartEvent(message_id=message_id, role="assistant")
-
-            # Check if this is confirm_changes (no steps) or function approval (has steps)
-            if not steps:
-                # This is confirm_changes for predictive state updates
-                if accepted:
-                    confirmation_message = strategy.on_state_confirmed()
-                else:
-                    confirmation_message = strategy.on_state_rejected()
-            elif accepted:
-                # User approved - execute the enabled steps (function approval flow)
-                confirmation_message = strategy.on_approval_accepted(steps)
-            else:
-                # User rejected
-                confirmation_message = strategy.on_approval_rejected(steps)
-
-            yield TextMessageContentEvent(
-                message_id=message_id,
-                delta=confirmation_message,
-            )
-
-            yield TextMessageEndEvent(message_id=message_id)
-
-            # Emit run finished
-            yield event_bridge.create_run_finished_event()
-
-        except json.JSONDecodeError:
-            logger.error(f"Failed to parse tool result: {tool_content_text}")
-            yield RunErrorEvent(message=f"Invalid tool result format: {tool_content_text[:100]}")
-            yield event_bridge.create_run_finished_event()
-
-
-class DefaultOrchestrator(Orchestrator):
-    """Standard agent execution (no special handling)."""
-
-    def can_handle(self, context: ExecutionContext) -> bool:
-        """Always returns True as this is the fallback orchestrator.
-
-        Args:
-            context: Execution context
-
-        Returns:
-            Always True
-        """
-        return True
-
-    def _create_initial_events(
-        self, event_bridge: "AgentFrameworkEventBridge", state_manager: "StateManager"
-    ) -> Sequence[BaseEvent]:
-        """Generate initial events for the run.
-
-        Args:
-            event_bridge: Event bridge for creating events
-        Returns:
-            Initial AG-UI events
-        """
-        events: list[BaseEvent] = [event_bridge.create_run_started_event()]
-
-        predict_event = state_manager.predict_state_event()
-        if predict_event:
-            events.append(predict_event)
-
-        snapshot_event = state_manager.initial_snapshot_event(event_bridge)
-        if snapshot_event:
-            events.append(snapshot_event)
-
-        return events
-
-    async def run(
-        self,
-        context: ExecutionContext,
-    ) -> AsyncGenerator[BaseEvent, None]:
-        """Standard agent run with event translation.
-
-        This implements the default agent execution flow using the event bridge
-        to translate Agent Framework events to AG-UI events.
-
-        Args:
-            context: Execution context
-
-        Yields:
-            AG-UI events
-        """
-        from ._events import AgentFrameworkEventBridge
-        from ._orchestration._state_manager import StateManager
-
-        logger.info(f"Starting default agent run for thread_id={context.thread_id}, run_id={context.run_id}")
-
-        response_format = None
-        if isinstance(context.agent, ChatAgent):
-            response_format = context.agent.default_options.get("response_format")
-        skip_text_content = response_format is not None
-
-        client_tools = convert_agui_tools_to_agent_framework(context.input_data.get("tools"))
-        approval_tool_name = select_approval_tool_name(client_tools)
-
-        state_manager = StateManager(
-            state_schema=context.config.state_schema,
-            predict_state_config=context.config.predict_state_config,
-            require_confirmation=context.config.require_confirmation,
-        )
-        current_state = state_manager.initialize(context.input_data.get("state"))
-
-        event_bridge = AgentFrameworkEventBridge(
-            run_id=context.run_id,
-            thread_id=context.thread_id,
-            predict_state_config=context.config.predict_state_config,
-            current_state=current_state,
-            skip_text_content=skip_text_content,
-            require_confirmation=context.config.require_confirmation,
-            approval_tool_name=approval_tool_name,
-        )
-
-        if context.config.use_service_thread:
-            thread = AgentThread(service_thread_id=context.supplied_thread_id)
-        else:
-            thread = AgentThread()
-
-        thread.metadata = {  # type: ignore[attr-defined]
-            "ag_ui_thread_id": context.thread_id,
-            "ag_ui_run_id": context.run_id,
-        }
-        if current_state:
-            thread.metadata["current_state"] = current_state  # type: ignore[attr-defined]
-
-        provider_messages = context.messages or []
-        snapshot_messages = context.snapshot_messages
-        if not provider_messages:
-            for event in self._create_initial_events(event_bridge, state_manager):
-                yield event
-            logger.warning("No messages provided in AG-UI input")
-            yield event_bridge.create_run_finished_event()
-            return
-
-        logger.info(f"Received {len(provider_messages)} provider messages from client")
-        for i, msg in enumerate(provider_messages):
-            role = get_role_value(msg)
-            msg_id = getattr(msg, "message_id", None)
-            logger.info(f"  Message {i}: role={role}, id={msg_id}")
-            if hasattr(msg, "contents") and msg.contents:
-                for j, content in enumerate(msg.contents):
-                    content_type = type(content).__name__
-                    if isinstance(content, TextContent):
-                        logger.debug("    Content %s: %s - text_length=%s", j, content_type, len(content.text))
-                    elif isinstance(content, FunctionCallContent):
-                        arg_length = len(str(content.arguments)) if content.arguments else 0
-                        logger.debug(
-                            "    Content %s: %s - %s args_length=%s", j, content_type, content.name, arg_length
-                        )
-                    elif isinstance(content, FunctionResultContent):
-                        result_preview = type(content.result).__name__ if content.result is not None else "None"
-                        logger.debug(
-                            "    Content %s: %s - call_id=%s, result_type=%s",
-                            j,
-                            content_type,
-                            content.call_id,
-                            result_preview,
-                        )
-                    else:
-                        logger.debug(f"    Content {j}: {content_type}")
-
-        pending_tool_calls: list[dict[str, Any]] = []
-        tool_calls_by_id: dict[str, dict[str, Any]] = {}
-        tool_results: list[dict[str, Any]] = []
-        tool_calls_ended: set[str] = set()
-        messages_snapshot_emitted = False
-        accumulated_text_content = ""
-        active_message_id: str | None = None
-
-        # Check for FunctionApprovalResponseContent and emit updated state snapshot
-        # This ensures the UI shows the approved state (e.g., 2 steps) not the original (3 steps)
-        for snapshot_evt in collect_approved_state_snapshots(
-            provider_messages,
-            context.config.predict_state_config,
-            current_state,
-            event_bridge,
-        ):
-            yield snapshot_evt
-
-        messages_to_run = select_messages_to_run(provider_messages, state_manager)
-
-        logger.info(f"[TOOLS] Client sent {len(client_tools) if client_tools else 0} tools")
-        if client_tools:
-            for tool in client_tools:
-                tool_name = getattr(tool, "name", "unknown")
-                declaration_only = getattr(tool, "declaration_only", None)
-                logger.info(f"[TOOLS]   - Client tool: {tool_name}, declaration_only={declaration_only}")
-
-        server_tools = collect_server_tools(context.agent)
-        register_additional_client_tools(context.agent, client_tools)
-        tools_param = merge_tools(server_tools, client_tools)
-
-        collect_updates = response_format is not None
-        all_updates: list[Any] | None = [] if collect_updates else None
-        update_count = 0
-        # Prepare metadata for chat client (Azure requires string values)
-        safe_metadata = build_safe_metadata(getattr(thread, "metadata", None))
-
-        run_kwargs: dict[str, Any] = {
-            "thread": thread,
-            "tools": tools_param,
-            "options": {"metadata": safe_metadata},
-        }
-        if safe_metadata:
-            run_kwargs["options"]["store"] = True
-
-        async def _resolve_approval_responses(
-            messages: list[Any],
-            tools_for_execution: list[Any],
-        ) -> None:
-            fcc_todo = _collect_approval_responses(messages)
-            if not fcc_todo:
-                return
-
-            approved_responses = [resp for resp in fcc_todo.values() if resp.approved]
-            approved_function_results: list[Any] = []
-            if approved_responses and tools_for_execution:
-                chat_client = getattr(context.agent, "chat_client", None)
-                config = (
-                    getattr(chat_client, "function_invocation_configuration", None) or FunctionInvocationConfiguration()
-                )
-                middleware_pipeline = extract_and_merge_function_middleware(chat_client, run_kwargs)
-                try:
-                    results, _ = await _try_execute_function_calls(
-                        custom_args=run_kwargs,
-                        attempt_idx=0,
-                        function_calls=approved_responses,
-                        tools=tools_for_execution,
-                        middleware_pipeline=middleware_pipeline,
-                        config=config,
-                    )
-                    approved_function_results = list(results)
-                except Exception:
-                    logger.error("Failed to execute approved tool calls; injecting error results.")
-                    approved_function_results = []
-
-            normalized_results: list[FunctionResultContent] = []
-            for idx, approval in enumerate(approved_responses):
-                if idx < len(approved_function_results) and isinstance(
-                    approved_function_results[idx], FunctionResultContent
-                ):
-                    normalized_results.append(approved_function_results[idx])
-                    continue
-                call_id = approval.function_call.call_id or approval.id
-                normalized_results.append(
-                    FunctionResultContent(call_id=call_id, result="Error: Tool call invocation failed.")
-                )
-
-            _replace_approval_contents_with_results(messages, fcc_todo, normalized_results)  # type: ignore
-
-        def _should_emit_tool_snapshot(tool_name: str | None) -> bool:
-            if not pending_tool_calls or not tool_results:
-                return False
-            if tool_name and context.config.predict_state_config and not context.config.require_confirmation:
-                for config in context.config.predict_state_config.values():
-                    if config["tool"] == tool_name:
-                        logger.info(
-                            f"Skipping intermediate MessagesSnapshotEvent for predictive tool '{tool_name}' "
-                            " - delaying until summary"
-                        )
-                        return False
-            return True
-
-        def _build_messages_snapshot(tool_message_id: str | None = None) -> MessagesSnapshotEvent:
-            has_text_content = bool(accumulated_text_content)
-            all_messages = snapshot_messages.copy()
-
-            if pending_tool_calls:
-                if tool_message_id and not has_text_content:
-                    tool_call_message_id = tool_message_id
-                else:
-                    tool_call_message_id = (
-                        active_message_id if not has_text_content and active_message_id else generate_event_id()
-                    )
-                tool_call_message = {
-                    "id": tool_call_message_id,
-                    "role": "assistant",
-                    "tool_calls": pending_tool_calls.copy(),
-                }
-                all_messages.append(tool_call_message)
-
-            all_messages.extend(tool_results)
-
-            if has_text_content and active_message_id:
-                assistant_text_message = {
-                    "id": active_message_id,
-                    "role": "assistant",
-                    "content": accumulated_text_content,
-                }
-                all_messages.append(assistant_text_message)
-
-            return MessagesSnapshotEvent(
-                messages=all_messages,  # type: ignore[arg-type]
-            )
-
-        # Use tools_param if available (includes client tools), otherwise fall back to server_tools
-        # This ensures both server tools AND client tools can be executed after approval
-        tools_for_approval = tools_param if tools_param is not None else server_tools
-        latest_approval = latest_approval_response(messages_to_run)
-        await _resolve_approval_responses(messages_to_run, tools_for_approval)
-
-        if latest_approval and is_step_based_approval(latest_approval, context.config.predict_state_config):
-            from ._confirmation_strategies import DefaultConfirmationStrategy
-
-            strategy = context.confirmation_strategy
-            if strategy is None:
-                strategy = DefaultConfirmationStrategy()
-
-            steps = approval_steps(latest_approval)
-            if steps:
-                if latest_approval.approved:
-                    confirmation_message = strategy.on_approval_accepted(steps)
-                else:
-                    confirmation_message = strategy.on_approval_rejected(steps)
-            else:
-                if latest_approval.approved:
-                    confirmation_message = strategy.on_state_confirmed()
-                else:
-                    confirmation_message = strategy.on_state_rejected()
-
-            message_id = generate_event_id()
-            for event in self._create_initial_events(event_bridge, state_manager):
-                yield event
-            yield TextMessageStartEvent(message_id=message_id, role="assistant")
-            yield TextMessageContentEvent(message_id=message_id, delta=confirmation_message)
-            yield TextMessageEndEvent(message_id=message_id)
-            yield event_bridge.create_run_finished_event()
-            return
-
-        should_recreate_event_bridge = False
-        async for update in context.agent.run_stream(messages_to_run, **run_kwargs):
-            conv_id = get_conversation_id_from_update(update)
-            if conv_id and conv_id != context.thread_id:
-                context.update_thread_id(conv_id)
-                should_recreate_event_bridge = True
-
-            if update.response_id and update.response_id != context.run_id:
-                context.update_run_id(update.response_id)
-                should_recreate_event_bridge = True
-
-            if should_recreate_event_bridge:
-                event_bridge = AgentFrameworkEventBridge(
-                    run_id=context.run_id,
-                    thread_id=context.thread_id,
-                    predict_state_config=context.config.predict_state_config,
-                    current_state=current_state,
-                    skip_text_content=skip_text_content,
-                    require_confirmation=context.config.require_confirmation,
-                    approval_tool_name=approval_tool_name,
-                )
-                should_recreate_event_bridge = False
-
-            if update_count == 0:
-                for event in self._create_initial_events(event_bridge, state_manager):
-                    yield event
-
-            update_count += 1
-            logger.info(f"[STREAM] Received update #{update_count} from agent")
-            if all_updates is not None:
-                all_updates.append(update)
-            if event_bridge.current_message_id is None and update.contents:
-                has_tool_call = any(isinstance(content, FunctionCallContent) for content in update.contents)
-                has_text = any(isinstance(content, TextContent) for content in update.contents)
-                if has_tool_call and not has_text:
-                    tool_message_id = generate_event_id()
-                    event_bridge.current_message_id = tool_message_id
-                    active_message_id = tool_message_id
-                    accumulated_text_content = ""
-                    logger.info(
-                        "[STREAM] Emitting TextMessageStartEvent for tool-only response message_id=%s",
-                        tool_message_id,
-                    )
-                    yield TextMessageStartEvent(message_id=tool_message_id, role="assistant")
-            events = await event_bridge.from_agent_run_update(update)
-            logger.info(f"[STREAM] Update #{update_count} produced {len(events)} events")
-            for event in events:
-                if isinstance(event, TextMessageStartEvent):
-                    active_message_id = event.message_id
-                    accumulated_text_content = ""
-                elif isinstance(event, TextMessageContentEvent):
-                    accumulated_text_content += event.delta
-                elif isinstance(event, ToolCallStartEvent):
-                    tool_call_entry = ensure_tool_call_entry(event.tool_call_id, tool_calls_by_id, pending_tool_calls)
-                    tool_call_entry["function"]["name"] = event.tool_call_name
-                elif isinstance(event, ToolCallArgsEvent):
-                    tool_call_entry = ensure_tool_call_entry(event.tool_call_id, tool_calls_by_id, pending_tool_calls)
-                    tool_call_entry["function"]["arguments"] += event.delta
-                elif isinstance(event, ToolCallEndEvent):
-                    tool_calls_ended.add(event.tool_call_id)
-                elif isinstance(event, ToolCallResultEvent):
-                    tool_results.append(
-                        {
-                            "id": event.message_id,
-                            "role": "tool",
-                            "toolCallId": event.tool_call_id,
-                            "content": event.content,
-                        }
-                    )
-                logger.info(f"[STREAM] Yielding event: {type(event).__name__}")
-                yield event
-                if isinstance(event, ToolCallResultEvent):
-                    tool_name = tool_name_for_call_id(tool_calls_by_id, event.tool_call_id)
-                    if _should_emit_tool_snapshot(tool_name):
-                        messages_snapshot_emitted = True
-                        messages_snapshot = _build_messages_snapshot()
-                        logger.info(f"[STREAM] Yielding event: {type(messages_snapshot).__name__}")
-                        yield messages_snapshot
-                elif isinstance(event, ToolCallEndEvent):
-                    tool_name = tool_name_for_call_id(tool_calls_by_id, event.tool_call_id)
-                    if tool_name == "confirm_changes":
-                        messages_snapshot_emitted = True
-                        messages_snapshot = _build_messages_snapshot()
-                        logger.info(f"[STREAM] Yielding event: {type(messages_snapshot).__name__}")
-                        yield messages_snapshot
-
-        logger.info(f"[STREAM] Agent stream completed. Total updates: {update_count}")
-
-        if event_bridge.should_stop_after_confirm:
-            logger.info("Stopping run - waiting for user approval/confirmation response")
-            if event_bridge.current_message_id:
-                logger.info(f"[CONFIRM] Emitting TextMessageEndEvent for message_id={event_bridge.current_message_id}")
-                yield event_bridge.create_message_end_event(event_bridge.current_message_id)
-                event_bridge.current_message_id = None
-            yield event_bridge.create_run_finished_event()
-            return
-
-        if pending_tool_calls:
-            pending_without_end = [tc for tc in pending_tool_calls if tc.get("id") not in tool_calls_ended]
-            if pending_without_end:
-                logger.info(
-                    "Found %s pending tool calls without end event - emitting ToolCallEndEvent",
-                    len(pending_without_end),
-                )
-                for tool_call in pending_without_end:
-                    tool_call_id = tool_call.get("id")
-                    if tool_call_id:
-                        end_event = ToolCallEndEvent(tool_call_id=tool_call_id)
-                        logger.info(f"Emitting ToolCallEndEvent for declaration-only tool call '{tool_call_id}'")
-                        yield end_event
-
-        if response_format and all_updates:
-            from agent_framework import AgentResponse
-            from pydantic import BaseModel
-
-            logger.info(f"Processing structured output, update count: {len(all_updates)}")
-            final_response = AgentResponse.from_agent_run_response_updates(
-                all_updates, output_format_type=response_format
-            )
-
-            if final_response.value and isinstance(final_response.value, BaseModel):
-                response_dict = final_response.value.model_dump(mode="json", exclude_none=True)
-                logger.info(f"Received structured output keys: {list(response_dict.keys())}")
-
-                state_updates = state_manager.extract_state_updates(response_dict)
-                if state_updates:
-                    state_manager.apply_state_updates(state_updates)
-                    state_snapshot = event_bridge.create_state_snapshot_event(current_state)
-                    yield state_snapshot
-                    logger.info(f"Emitted StateSnapshotEvent with updates: {list(state_updates.keys())}")
-
-                if "message" in response_dict and response_dict["message"]:
-                    message_id = generate_event_id()
-                    yield TextMessageStartEvent(message_id=message_id, role="assistant")
-                    yield TextMessageContentEvent(message_id=message_id, delta=response_dict["message"])
-                    yield TextMessageEndEvent(message_id=message_id)
-                    logger.info(f"Emitted conversational message with length={len(response_dict['message'])}")
-
-        if all_updates is not None and len(all_updates) == 0:
-            logger.info("No updates received from agent - emitting initial events")
-            for event in self._create_initial_events(event_bridge, state_manager):
-                yield event
-
-        logger.info(f"[FINALIZE] Checking for unclosed message. current_message_id={event_bridge.current_message_id}")
-        if event_bridge.current_message_id:
-            logger.info(f"[FINALIZE] Emitting TextMessageEndEvent for message_id={event_bridge.current_message_id}")
-            yield event_bridge.create_message_end_event(event_bridge.current_message_id)
-
-            messages_snapshot = _build_messages_snapshot(tool_message_id=event_bridge.current_message_id)
-            messages_snapshot_emitted = True
-            logger.info(
-                f"[FINALIZE] Emitting MessagesSnapshotEvent with {len(messages_snapshot.messages)} messages "
-                f"(text content length: {len(accumulated_text_content)})"
-            )
-            yield messages_snapshot
-        else:
-            logger.info("[FINALIZE] No current_message_id - skipping TextMessageEndEvent")
-            if not messages_snapshot_emitted and (pending_tool_calls or tool_results):
-                messages_snapshot = _build_messages_snapshot()
-                messages_snapshot_emitted = True
-                logger.info(
-                    f"[FINALIZE] Emitting MessagesSnapshotEvent with {len(messages_snapshot.messages)} messages"
-                )
-                yield messages_snapshot
-
-        logger.info("[FINALIZE] Emitting RUN_FINISHED event")
-        yield event_bridge.create_run_finished_event()
-        logger.info(f"Completed agent run for thread_id={context.thread_id}, run_id={context.run_id}")
-
-
-__all__ = [
-    "Orchestrator",
-    "ExecutionContext",
-    "HumanInTheLoopOrchestrator",
-    "DefaultOrchestrator",
-]
@@ -0,0 +1,963 @@
+# Copyright (c) Microsoft. All rights reserved.
+
+"""Simplified AG-UI orchestration - single linear flow."""
+
+import json
+import logging
+import uuid
+from dataclasses import dataclass, field
+from typing import TYPE_CHECKING, Any
+
+from ag_ui.core import (
+    BaseEvent,
+    CustomEvent,
+    MessagesSnapshotEvent,
+    RunFinishedEvent,
+    RunStartedEvent,
+    StateSnapshotEvent,
+    TextMessageContentEvent,
+    TextMessageEndEvent,
+    TextMessageStartEvent,
+    ToolCallArgsEvent,
+    ToolCallEndEvent,
+    ToolCallResultEvent,
+    ToolCallStartEvent,
+)
+from agent_framework import (
+    AgentProtocol,
+    AgentThread,
+    ChatMessage,
+    Content,
+    prepare_function_call_results,
+)
+from agent_framework._middleware import extract_and_merge_function_middleware
+from agent_framework._tools import (
+    FunctionInvocationConfiguration,
+    _collect_approval_responses,  # type: ignore
+    _replace_approval_contents_with_results,  # type: ignore
+    _try_execute_function_calls,  # type: ignore
+)
+
+from ._message_adapters import normalize_agui_input_messages
+from ._orchestration._predictive_state import PredictiveStateHandler
+from ._orchestration._tooling import collect_server_tools, merge_tools, register_additional_client_tools
+from ._utils import (
+    convert_agui_tools_to_agent_framework,
+    generate_event_id,
+    get_conversation_id_from_update,
+    make_json_safe,
+)
+
+if TYPE_CHECKING:
+    from collections.abc import AsyncGenerator
+
+    from ._agent import AgentConfig
+
+logger = logging.getLogger(__name__)
+
+# Keys that are internal to AG-UI orchestration and should not be passed to chat clients
+AG_UI_INTERNAL_METADATA_KEYS = {"ag_ui_thread_id", "ag_ui_run_id", "current_state"}
+
+
+def _build_safe_metadata(thread_metadata: dict[str, Any] | None) -> dict[str, Any]:
+    """Build metadata dict with truncated string values for Azure compatibility.
+
+    Azure has a 512 character limit per metadata value.
+
+    Args:
+        thread_metadata: Raw metadata dict
+
+    Returns:
+        Metadata with string values truncated to 512 chars
+    """
+    if not thread_metadata:
+        return {}
+    safe_metadata: dict[str, Any] = {}
+    for key, value in thread_metadata.items():
+        value_str = value if isinstance(value, str) else json.dumps(value)
+        if len(value_str) > 512:
+            value_str = value_str[:512]
+        safe_metadata[key] = value_str
+    return safe_metadata
+
+
+def _has_only_tool_calls(contents: list[Any]) -> bool:
+    """Check if contents have only tool calls (no text).
+
+    Args:
+        contents: List of content items
+
+    Returns:
+        True if there are tool calls but no text content
+    """
+    has_tool_call = any(getattr(c, "type", None) == "function_call" for c in contents)
+    has_text = any(getattr(c, "type", None) == "text" and getattr(c, "text", None) for c in contents)
+    return has_tool_call and not has_text
+
+
+def _should_suppress_intermediate_snapshot(
+    tool_name: str | None,
+    predict_state_config: dict[str, dict[str, str]] | None,
+    require_confirmation: bool,
+) -> bool:
+    """Check if intermediate MessagesSnapshotEvent should be suppressed for this tool.
+
+    For predictive tools without confirmation, we delay the snapshot until the end.
+
+    Args:
+        tool_name: Name of the tool that just completed
+        predict_state_config: Predictive state configuration
+        require_confirmation: Whether confirmation is required
+
+    Returns:
+        True if snapshot should be suppressed
+    """
+    if not tool_name or not predict_state_config:
+        return False
+    # Only suppress when confirmation is disabled
+    if require_confirmation:
+        return False
+    # Check if this tool is a predictive tool
+    for config in predict_state_config.values():
+        if config["tool"] == tool_name:
+            logger.info(f"Suppressing intermediate MessagesSnapshotEvent for predictive tool '{tool_name}'")
+            return True
+    return False
+
+
+def _extract_approved_state_updates(
+    messages: list[Any],
+    predictive_handler: PredictiveStateHandler | None,
+) -> dict[str, Any]:
+    """Extract state updates from function_approval_response content.
+
+    This emits StateSnapshotEvent for approved state-changing tools before running agent.
+
+    Args:
+        messages: List of messages to scan
+        predictive_handler: Predictive state handler
+
+    Returns:
+        Dict of state updates to apply
+    """
+    if not predictive_handler:
+        return {}
+
+    updates: dict[str, Any] = {}
+    for msg in messages:
+        for content in msg.contents:
+            if getattr(content, "type", None) != "function_approval_response":
+                continue
+            if not getattr(content, "approved", False) or not getattr(content, "function_call", None):
+                continue
+            parsed_args = content.function_call.parse_arguments()
+            result = predictive_handler.extract_state_value(content.function_call.name, parsed_args)
+            if result:
+                state_key, state_value = result
+                updates[state_key] = state_value
+                logger.info(f"Found approved state update for key '{state_key}'")
+    return updates
+
+
+@dataclass
+class FlowState:
+    """Minimal explicit state for a single AG-UI run."""
+
+    message_id: str | None = None  # Current text message being streamed
+    tool_call_id: str | None = None  # Current tool call being streamed
+    tool_call_name: str | None = None  # Name of current tool call
+    waiting_for_approval: bool = False  # Stop after approval request
+    current_state: dict[str, Any] = field(default_factory=dict)  # Shared state
+    accumulated_text: str = ""  # For MessagesSnapshotEvent
+    pending_tool_calls: list[dict[str, Any]] = field(default_factory=list)  # For MessagesSnapshotEvent
+    tool_calls_by_id: dict[str, dict[str, Any]] = field(default_factory=dict)
+    tool_results: list[dict[str, Any]] = field(default_factory=list)
+    tool_calls_ended: set[str] = field(default_factory=set)  # Track which tool calls have been ended
+
+    def get_tool_name(self, call_id: str | None) -> str | None:
+        """Get tool name by call ID."""
+        if not call_id or call_id not in self.tool_calls_by_id:
+            return None
+        name = self.tool_calls_by_id[call_id]["function"].get("name")
+        return str(name) if name else None
+
+    def get_pending_without_end(self) -> list[dict[str, Any]]:
+        """Get tool calls that started but never received an end event (declaration-only)."""
+        return [tc for tc in self.pending_tool_calls if tc.get("id") not in self.tool_calls_ended]
+
+
+def _create_state_context_message(
+    current_state: dict[str, Any],
+    state_schema: dict[str, Any],
+) -> ChatMessage | None:
+    """Create a system message with current state context.
+
+    This injects the current state into the conversation so the model
+    knows what state exists and can make informed updates.
+
+    Args:
+        current_state: The current state to inject
+        state_schema: The state schema (used to determine if injection is needed)
+
+    Returns:
+        ChatMessage with state context, or None if not needed
+    """
+    if not current_state or not state_schema:
+        return None
+
+    state_json = json.dumps(current_state, indent=2)
+    return ChatMessage(
+        role="system",
+        contents=[
+            Content.from_text(
+                text=(
+                    "Current state of the application:\n"
+                    f"{state_json}\n\n"
+                    "When modifying state, you MUST include ALL existing data plus your changes.\n"
+                    "For example, if adding one new item to a list, include ALL existing items PLUS the new item.\n"
+                    "Never replace existing data - always preserve and append or merge."
+                )
+            )
+        ],
+    )
+
+
+def _inject_state_context(
+    messages: list[ChatMessage],
+    current_state: dict[str, Any],
+    state_schema: dict[str, Any],
+) -> list[ChatMessage]:
+    """Inject state context message into messages if appropriate.
+
+    The state context is injected before the last user message to give
+    the model visibility into the current application state.
+
+    Args:
+        messages: The messages to potentially inject into
+        current_state: The current state
+        state_schema: The state schema
+
+    Returns:
+        Messages with state context injected if appropriate
+    """
+    state_msg = _create_state_context_message(current_state, state_schema)
+    if not state_msg:
+        return messages
+
+    # Check if the last message is from a user (new user turn)
+    if not messages:
+        return messages
+
+    from ._utils import get_role_value
+
+    last_role = get_role_value(messages[-1])
+    if last_role != "user":
+        return messages
+
+    # Always inject state context if state is provided
+    # This ensures UI state changes are visible to the model
+
+    # Insert state context before the last user message
+    result = list(messages[:-1])
+    result.append(state_msg)
+    result.append(messages[-1])
+    return result
+
+
+def _emit_text(content: Content, flow: FlowState, skip_text: bool = False) -> list[BaseEvent]:
+    """Emit TextMessage events for TextContent."""
+    if not content.text:
+        return []
+
+    # Skip if we're in structured output mode or waiting for approval
+    if skip_text or flow.waiting_for_approval:
+        return []
+
+    events: list[BaseEvent] = []
+    if not flow.message_id:
+        flow.message_id = generate_event_id()
+        events.append(TextMessageStartEvent(message_id=flow.message_id, role="assistant"))
+
+    events.append(TextMessageContentEvent(message_id=flow.message_id, delta=content.text))
+    flow.accumulated_text += content.text
+    return events
+
+
+def _emit_tool_call(
+    content: Content,
+    flow: FlowState,
+    predictive_handler: PredictiveStateHandler | None = None,
+) -> list[BaseEvent]:
+    """Emit ToolCall events for FunctionCallContent."""
+    events: list[BaseEvent] = []
+
+    tool_call_id = content.call_id or flow.tool_call_id or generate_event_id()
+
+    # Emit start event when we have a new tool call
+    if content.name and tool_call_id != flow.tool_call_id:
+        flow.tool_call_id = tool_call_id
+        flow.tool_call_name = content.name
+        if predictive_handler:
+            predictive_handler.reset_streaming()
+
+        events.append(
+            ToolCallStartEvent(
+                tool_call_id=tool_call_id,
+                tool_call_name=content.name,
+                parent_message_id=flow.message_id,
+            )
+        )
+
+        # Track for MessagesSnapshotEvent
+        tool_entry = {
+            "id": tool_call_id,
+            "type": "function",
+            "function": {"name": content.name, "arguments": ""},
+        }
+        flow.pending_tool_calls.append(tool_entry)
+        flow.tool_calls_by_id[tool_call_id] = tool_entry
+
+    elif tool_call_id:
+        flow.tool_call_id = tool_call_id
+
+    # Emit args if present
+    if content.arguments:
+        delta = (
+            content.arguments if isinstance(content.arguments, str) else json.dumps(make_json_safe(content.arguments))
+        )
+        events.append(ToolCallArgsEvent(tool_call_id=tool_call_id, delta=delta))
+
+        # Track args for MessagesSnapshotEvent
+        if tool_call_id in flow.tool_calls_by_id:
+            flow.tool_calls_by_id[tool_call_id]["function"]["arguments"] += delta
+
+        # Emit predictive state deltas
+        if predictive_handler and flow.tool_call_name:
+            delta_events = predictive_handler.emit_streaming_deltas(flow.tool_call_name, delta)
+            events.extend(delta_events)
+
+    return events
+
+
+def _emit_tool_result(
+    content: Content,
+    flow: FlowState,
+    predictive_handler: PredictiveStateHandler | None = None,
+) -> list[BaseEvent]:
+    """Emit ToolCallResult events for FunctionResultContent."""
+    events: list[BaseEvent] = []
+
+    # Cannot emit tool result without a call_id to associate it with
+    if not content.call_id:
+        return events
+
+    events.append(ToolCallEndEvent(tool_call_id=content.call_id))
+    flow.tool_calls_ended.add(content.call_id)  # Track ended tool calls
+
+    result_content = prepare_function_call_results(content.result)
+    message_id = generate_event_id()
+    events.append(
+        ToolCallResultEvent(
+            message_id=message_id,
+            tool_call_id=content.call_id,
+            content=result_content,
+            role="tool",
+        )
+    )
+
+    # Track for MessagesSnapshotEvent
+    flow.tool_results.append(
+        {
+            "id": message_id,
+            "role": "tool",
+            "toolCallId": content.call_id,
+            "content": result_content,
+        }
+    )
+
+    # Apply predictive state updates and emit snapshot
+    if predictive_handler:
+        predictive_handler.apply_pending_updates()
+        if flow.current_state:
+            events.append(StateSnapshotEvent(snapshot=flow.current_state))
+
+    # Reset tool tracking and message context
+    # After tool result, any subsequent text should start a new message
+    flow.tool_call_id = None
+    flow.tool_call_name = None
+    flow.message_id = None  # Reset so next text content starts a new message
+
+    return events
+
+
+def _emit_approval_request(
+    content: Content,
+    flow: FlowState,
+    predictive_handler: PredictiveStateHandler | None = None,
+    require_confirmation: bool = True,
+) -> list[BaseEvent]:
+    """Emit events for function approval request."""
+    events: list[BaseEvent] = []
+
+    # function_call is required for approval requests - skip if missing
+    func_call = content.function_call
+    if not func_call:
+        logger.warning("Approval request content missing function_call, skipping")
+        return events
+
+    func_name = func_call.name or ""
+    func_call_id = func_call.call_id
+
+    # Extract state from function arguments if predictive
+    if predictive_handler and func_name:
+        parsed_args = func_call.parse_arguments()
+        result = predictive_handler.extract_state_value(func_name, parsed_args)
+        if result:
+            state_key, state_value = result
+            flow.current_state[state_key] = state_value
+            events.append(StateSnapshotEvent(snapshot=flow.current_state))
+
+    # End the original tool call
+    if func_call_id:
+        events.append(ToolCallEndEvent(tool_call_id=func_call_id))
+        flow.tool_calls_ended.add(func_call_id)  # Track ended tool calls
+
+    # Emit custom event for UI
+    events.append(
+        CustomEvent(
+            name="function_approval_request",
+            value={
+                "id": content.id,
+                "function_call": {
+                    "call_id": func_call_id,
+                    "name": func_name,
+                    "arguments": make_json_safe(func_call.parse_arguments()),
+                },
+            },
+        )
+    )
+
+    # Emit confirm_changes tool call for UI compatibility
+    # The complete sequence (Start -> Args -> End) signals the UI to show the confirmation dialog
+    if require_confirmation:
+        confirm_id = generate_event_id()
+        events.append(
+            ToolCallStartEvent(
+                tool_call_id=confirm_id,
+                tool_call_name="confirm_changes",
+                parent_message_id=flow.message_id,
+            )
+        )
+        args = {
+            "function_name": func_name,
+            "function_call_id": func_call_id,
+            "function_arguments": make_json_safe(func_call.parse_arguments()) or {},
+            "steps": [{"description": f"Execute {func_name}", "status": "enabled"}],
+        }
+        events.append(ToolCallArgsEvent(tool_call_id=confirm_id, delta=json.dumps(args)))
+        events.append(ToolCallEndEvent(tool_call_id=confirm_id))
+
+    flow.waiting_for_approval = True
+    return events
+
+
+def _emit_content(
+    content: Any,
+    flow: FlowState,
+    predictive_handler: PredictiveStateHandler | None = None,
+    skip_text: bool = False,
+    require_confirmation: bool = True,
+) -> list[BaseEvent]:
+    """Emit appropriate events for any content type."""
+    content_type = getattr(content, "type", None)
+    if content_type == "text":
+        return _emit_text(content, flow, skip_text)
+    elif content_type == "function_call":
+        return _emit_tool_call(content, flow, predictive_handler)
+    elif content_type == "function_result":
+        return _emit_tool_result(content, flow, predictive_handler)
+    elif content_type == "function_approval_request":
+        return _emit_approval_request(content, flow, predictive_handler, require_confirmation)
+    return []
+
+
+def _is_confirm_changes_response(messages: list[Any]) -> bool:
+    """Check if the last message is a confirm_changes tool result (state confirmation flow).
+
+    This returns True for confirm_changes flows where we emit a confirmation message
+    and stop. The key indicator is the presence of a 'steps' key in the tool result
+    (even if empty), combined with 'accepted' boolean.
+    """
+    if not messages:
+        return False
+    last = messages[-1]
+    if not last.additional_properties.get("is_tool_result", False):
+        return False
+
+    # Parse the content to check if it has the confirm_changes structure
+    for content in last.contents:
+        if getattr(content, "type", None) == "text":
+            try:
+                result = json.loads(content.text)
+                # confirm_changes results have 'accepted' and 'steps' keys
+                if "accepted" in result and "steps" in result:
+                    return True
+            except json.JSONDecodeError:
+                # Content is not valid JSON; continue checking other content items
+                logger.debug("Failed to parse confirm_changes tool result as JSON; treating as non-confirmation.")
+    return False
+
+
+def _handle_step_based_approval(messages: list[Any]) -> list[BaseEvent]:
+    """Handle step-based approval response and emit confirmation message."""
+    events: list[BaseEvent] = []
+    last = messages[-1]
+
+    # Parse the approval content
+    approval_text = ""
+    for content in last.contents:
+        if getattr(content, "type", None) == "text":
+            approval_text = content.text
+            break
+
+    try:
+        result = json.loads(approval_text)
+        accepted = result.get("accepted", False)
+        steps = result.get("steps", [])
+
+        if accepted:
+            # Generate acceptance message with step descriptions
+            enabled_steps = [s for s in steps if s.get("status") == "enabled"]
+            if enabled_steps:
+                message_parts = [f"Executing {len(enabled_steps)} approved steps:\n\n"]
+                for i, step in enumerate(enabled_steps, 1):
+                    message_parts.append(f"{i}. {step.get('description', 'Step')}\n")
+                message_parts.append("\nAll steps completed successfully!")
+                message = "".join(message_parts)
+            else:
+                message = "Changes confirmed and applied successfully!"
+        else:
+            # Rejection message
+            message = "No problem! What would you like me to change about the plan?"
+    except json.JSONDecodeError:
+        message = "Acknowledged."
+
+    message_id = generate_event_id()
+    events.append(TextMessageStartEvent(message_id=message_id, role="assistant"))
+    events.append(TextMessageContentEvent(message_id=message_id, delta=message))
+    events.append(TextMessageEndEvent(message_id=message_id))
+
+    return events
+
+
+async def _resolve_approval_responses(
+    messages: list[Any],
+    tools: list[Any],
+    agent: AgentProtocol,
+    run_kwargs: dict[str, Any],
+) -> None:
+    """Execute approved function calls and replace approval content with results.
+
+    This modifies the messages list in place, replacing FunctionApprovalResponseContent
+    with FunctionResultContent containing the actual tool execution result.
+
+    Args:
+        messages: List of messages (will be modified in place)
+        tools: List of available tools
+        agent: The agent instance (to get chat_client and config)
+        run_kwargs: Kwargs for tool execution
+    """
+    fcc_todo = _collect_approval_responses(messages)
+    if not fcc_todo:
+        return
+
+    approved_responses = [resp for resp in fcc_todo.values() if resp.approved]
+    rejected_responses = [resp for resp in fcc_todo.values() if not resp.approved]
+    approved_function_results: list[Any] = []
+
+    # Execute approved tool calls
+    if approved_responses and tools:
+        chat_client = getattr(agent, "chat_client", None)
+        config = getattr(chat_client, "function_invocation_configuration", None) or FunctionInvocationConfiguration()
+        middleware_pipeline = extract_and_merge_function_middleware(chat_client, run_kwargs)
+        # Filter out AG-UI-specific kwargs that should not be passed to tool execution
+        tool_kwargs = {k: v for k, v in run_kwargs.items() if k != "options"}
+        try:
+            results, _ = await _try_execute_function_calls(
+                custom_args=tool_kwargs,
+                attempt_idx=0,
+                function_calls=approved_responses,
+                tools=tools,
+                middleware_pipeline=middleware_pipeline,
+                config=config,
+            )
+            approved_function_results = list(results)
+        except Exception as e:
+            logger.exception("Failed to execute approved tool calls; injecting error results: %s", e)
+            approved_function_results = []
+
+    # Build normalized results for approved responses
+    normalized_results: list[Content] = []
+    for idx, approval in enumerate(approved_responses):
+        if (
+            idx < len(approved_function_results)
+            and getattr(approved_function_results[idx], "type", None) == "function_result"
+        ):
+            normalized_results.append(approved_function_results[idx])
+            continue
+        # Get call_id from function_call if present, otherwise use approval.id
+        func_call = approval.function_call
+        call_id = (func_call.call_id if func_call else None) or approval.id or ""
+        normalized_results.append(
+            Content.from_function_result(call_id=call_id, result="Error: Tool call invocation failed.")
+        )
+
+    # Build rejection results
+    for rejection in rejected_responses:
+        func_call = rejection.function_call
+        call_id = (func_call.call_id if func_call else None) or rejection.id or ""
+        normalized_results.append(
+            Content.from_function_result(call_id=call_id, result="Error: Tool call invocation was rejected by user.")
+        )
+
+    _replace_approval_contents_with_results(messages, fcc_todo, normalized_results)  # type: ignore
+
+
+def _build_messages_snapshot(
+    flow: FlowState,
+    snapshot_messages: list[dict[str, Any]],
+) -> MessagesSnapshotEvent:
+    """Build MessagesSnapshotEvent from current flow state."""
+    all_messages = list(snapshot_messages)
+
+    # Add assistant message with tool calls
+    if flow.pending_tool_calls:
+        tool_call_message = {
+            "id": flow.message_id or generate_event_id(),
+            "role": "assistant",
+            "tool_calls": flow.pending_tool_calls.copy(),
+        }
+        if flow.accumulated_text:
+            tool_call_message["content"] = flow.accumulated_text
+        all_messages.append(tool_call_message)
+
+    # Add tool results
+    all_messages.extend(flow.tool_results)
+
+    # Add text-only assistant message if no tool calls
+    if flow.accumulated_text and not flow.pending_tool_calls:
+        all_messages.append(
+            {
+                "id": flow.message_id or generate_event_id(),
+                "role": "assistant",
+                "content": flow.accumulated_text,
+            }
+        )
+
+    return MessagesSnapshotEvent(messages=all_messages)  # type: ignore[arg-type]
+
+
+async def run_agent_stream(
+    input_data: dict[str, Any],
+    agent: AgentProtocol,
+    config: "AgentConfig",
+) -> "AsyncGenerator[BaseEvent, None]":
+    """Run agent and yield AG-UI events.
+
+    This is the single entry point for all AG-UI agent runs. It follows a simple
+    linear flow: RunStarted -> content events -> RunFinished.
+
+    Args:
+        input_data: AG-UI request data with messages, state, tools, etc.
+        agent: The Agent Framework agent to run
+        config: Agent configuration
+
+    Yields:
+        AG-UI events
+    """
+    # Parse IDs
+    thread_id = input_data.get("thread_id") or input_data.get("threadId") or str(uuid.uuid4())
+    run_id = input_data.get("run_id") or input_data.get("runId") or str(uuid.uuid4())
+
+    # Initialize flow state with schema defaults
+    flow = FlowState()
+    if input_data.get("state"):
+        flow.current_state = dict(input_data["state"])
+
+    # Apply schema defaults for missing state keys
+    if config.state_schema:
+        for key, schema in config.state_schema.items():
+            if key in flow.current_state:
+                continue
+            if isinstance(schema, dict) and schema.get("type") == "array":
+                flow.current_state[key] = []
+            else:
+                flow.current_state[key] = {}
+
+    # Initialize predictive state handler if configured
+    predictive_handler: PredictiveStateHandler | None = None
+    if config.predict_state_config:
+        predictive_handler = PredictiveStateHandler(
+            predict_state_config=config.predict_state_config,
+            current_state=flow.current_state,
+        )
+
+    # Normalize messages
+    raw_messages = input_data.get("messages", [])
+    messages, snapshot_messages = normalize_agui_input_messages(raw_messages)
+
+    # Check for structured output mode (skip text content)
+    skip_text = False
+    response_format = None
+    from agent_framework import ChatAgent
+
+    if isinstance(agent, ChatAgent):
+        response_format = agent.default_options.get("response_format")
+        skip_text = response_format is not None
+
+    # Handle empty messages (emit RunStarted immediately since no agent response)
+    if not messages:
+        logger.warning("No messages provided in AG-UI input")
+        yield RunStartedEvent(run_id=run_id, thread_id=thread_id)
+        yield RunFinishedEvent(run_id=run_id, thread_id=thread_id)
+        return
+
+    # Prepare tools
+    client_tools = convert_agui_tools_to_agent_framework(input_data.get("tools"))
+    server_tools = collect_server_tools(agent)
+    register_additional_client_tools(agent, client_tools)
+    tools = merge_tools(server_tools, client_tools)
+
+    # Create thread (with service thread support)
+    if config.use_service_thread:
+        supplied_thread_id = input_data.get("thread_id") or input_data.get("threadId")
+        thread = AgentThread(service_thread_id=supplied_thread_id)
+    else:
+        thread = AgentThread()
+
+    # Inject metadata for AG-UI orchestration (Feature #2: Azure-safe truncation)
+    base_metadata: dict[str, Any] = {
+        "ag_ui_thread_id": thread_id,
+        "ag_ui_run_id": run_id,
+    }
+    if flow.current_state:
+        base_metadata["current_state"] = flow.current_state
+    thread.metadata = _build_safe_metadata(base_metadata)  # type: ignore[attr-defined]
+
+    # Build run kwargs (Feature #6: Azure store flag when metadata present)
+    run_kwargs: dict[str, Any] = {"thread": thread}
+    if tools:
+        run_kwargs["tools"] = tools
+    # Filter out AG-UI internal metadata keys before passing to chat client
+    # These are used internally for orchestration and should not be sent to the LLM provider
+    client_metadata = {
+        k: v for k, v in (getattr(thread, "metadata", None) or {}).items() if k not in AG_UI_INTERNAL_METADATA_KEYS
+    }
+    safe_metadata = _build_safe_metadata(client_metadata) if client_metadata else {}
+    if safe_metadata:
+        run_kwargs["options"] = {"metadata": safe_metadata, "store": True}
+
+    # Resolve approval responses (execute approved tools, replace approvals with results)
+    # This must happen before running the agent so it sees the tool results
+    tools_for_execution = tools if tools is not None else server_tools
+    await _resolve_approval_responses(messages, tools_for_execution, agent, run_kwargs)
+
+    # Feature #3: Emit StateSnapshotEvent for approved state-changing tools before agent runs
+    approved_state_updates = _extract_approved_state_updates(messages, predictive_handler)
+    approved_state_snapshot_emitted = False
+    if approved_state_updates:
+        flow.current_state.update(approved_state_updates)
+        approved_state_snapshot_emitted = True
+
+    # Handle confirm_changes response (state confirmation flow - emit confirmation and stop)
+    if _is_confirm_changes_response(messages):
+        yield RunStartedEvent(run_id=run_id, thread_id=thread_id)
+        # Emit approved state snapshot before confirmation message
+        if approved_state_snapshot_emitted:
+            yield StateSnapshotEvent(snapshot=flow.current_state)
+        for event in _handle_step_based_approval(messages):
+            yield event
+        yield RunFinishedEvent(run_id=run_id, thread_id=thread_id)
+        return
+
+    # Inject state context message so the model knows current application state
+    # This is critical for shared state scenarios where the UI state needs to be visible
+    if config.state_schema and flow.current_state:
+        messages = _inject_state_context(messages, flow.current_state, config.state_schema)
+
+    # Stream from agent - emit RunStarted after first update to get service IDs
+    run_started_emitted = False
+    all_updates: list[Any] = []  # Collect for structured output processing
+    async for update in agent.run_stream(messages, **run_kwargs):
+        # Collect updates for structured output processing
+        if response_format is not None:
+            all_updates.append(update)
+
+        # Update IDs from service response on first update and emit RunStarted
+        if not run_started_emitted:
+            conv_id = get_conversation_id_from_update(update)
+            if conv_id:
+                thread_id = conv_id
+            if update.response_id:
+                run_id = update.response_id
+            # NOW emit RunStarted with proper IDs
+            yield RunStartedEvent(run_id=run_id, thread_id=thread_id)
+            # Emit PredictState custom event if configured
+            if config.predict_state_config:
+                predict_state_value = [
+                    {
+                        "state_key": state_key,
+                        "tool": cfg["tool"],
+                        "tool_argument": cfg["tool_argument"],
+                    }
+                    for state_key, cfg in config.predict_state_config.items()
+                ]
+                yield CustomEvent(name="PredictState", value=predict_state_value)
+            # Emit initial state snapshot only if we have both state_schema and state
+            if config.state_schema and flow.current_state:
+                yield StateSnapshotEvent(snapshot=flow.current_state)
+            run_started_emitted = True
+
+        # Feature #4: Detect tool-only messages (no text content)
+        # Emit TextMessageStartEvent to create message context for tool calls
+        if not flow.message_id and _has_only_tool_calls(update.contents):
+            flow.message_id = generate_event_id()
+            logger.info(f"Tool-only response detected, creating message_id={flow.message_id}")
+            yield TextMessageStartEvent(message_id=flow.message_id, role="assistant")
+
+        # Emit events for each content item
+        for content in update.contents:
+            for event in _emit_content(
+                content,
+                flow,
+                predictive_handler,
+                skip_text,
+                config.require_confirmation,
+            ):
+                yield event
+
+        # Stop if waiting for approval
+        if flow.waiting_for_approval:
+            break
+
+    # If no updates at all, still emit RunStarted
+    if not run_started_emitted:
+        yield RunStartedEvent(run_id=run_id, thread_id=thread_id)
+        if config.predict_state_config:
+            predict_state_value = [
+                {
+                    "state_key": state_key,
+                    "tool": cfg["tool"],
+                    "tool_argument": cfg["tool_argument"],
+                }
+                for state_key, cfg in config.predict_state_config.items()
+            ]
+            yield CustomEvent(name="PredictState", value=predict_state_value)
+        if config.state_schema and flow.current_state:
+            yield StateSnapshotEvent(snapshot=flow.current_state)
+
+    # Process structured output if response_format is set
+    if response_format is not None and all_updates:
+        from agent_framework import AgentResponse
+        from pydantic import BaseModel
+
+        logger.info(f"Processing structured output, update count: {len(all_updates)}")
+        final_response = AgentResponse.from_agent_run_response_updates(all_updates, output_format_type=response_format)
+
+        if final_response.value and isinstance(final_response.value, BaseModel):
+            response_dict = final_response.value.model_dump(mode="json", exclude_none=True)
+            logger.info(f"Received structured output keys: {list(response_dict.keys())}")
+
+            # Extract state updates - if no state_schema, all non-message fields are state
+            state_keys = (
+                set(config.state_schema.keys()) if config.state_schema else set(response_dict.keys()) - {"message"}
+            )
+            state_updates = {k: v for k, v in response_dict.items() if k in state_keys}
+
+            if state_updates:
+                flow.current_state.update(state_updates)
+                yield StateSnapshotEvent(snapshot=flow.current_state)
+                logger.info(f"Emitted StateSnapshotEvent with updates: {list(state_updates.keys())}")
+
+            # Emit message field as text if present
+            if "message" in response_dict and response_dict["message"]:
+                message_id = generate_event_id()
+                yield TextMessageStartEvent(message_id=message_id, role="assistant")
+                yield TextMessageContentEvent(message_id=message_id, delta=response_dict["message"])
+                yield TextMessageEndEvent(message_id=message_id)
+                logger.info(f"Emitted conversational message with length={len(response_dict['message'])}")
+
+    # Feature #1: Emit ToolCallEndEvent for declaration-only tools (tools without results)
+    pending_without_end = flow.get_pending_without_end()
+    if pending_without_end:
+        logger.info(f"Found {len(pending_without_end)} pending tool calls without end event")
+        for tool_call in pending_without_end:
+            tool_call_id = tool_call.get("id")
+            tool_name = tool_call.get("function", {}).get("name")
+            if tool_call_id:
+                logger.info(f"Emitting ToolCallEndEvent for declaration-only tool '{tool_call_id}'")
+                yield ToolCallEndEvent(tool_call_id=tool_call_id)
+
+                # For predictive tools with require_confirmation, emit confirm_changes
+                if config.require_confirmation and config.predict_state_config and tool_name:
+                    is_predictive_tool = any(cfg["tool"] == tool_name for cfg in config.predict_state_config.values())
+                    if is_predictive_tool:
+                        logger.info(f"Emitting confirm_changes for predictive tool '{tool_name}'")
+                        # Extract state value from tool arguments for StateSnapshot
+                        if predictive_handler:
+                            try:
+                                args_str = tool_call.get("function", {}).get("arguments", "{}")
+                                args = json.loads(args_str) if isinstance(args_str, str) else args_str
+                                result = predictive_handler.extract_state_value(tool_name, args)
+                                if result:
+                                    state_key, state_value = result
+                                    flow.current_state[state_key] = state_value
+                                    yield StateSnapshotEvent(snapshot=flow.current_state)
+                            except json.JSONDecodeError:
+                                # Ignore malformed JSON in tool arguments for predictive state;
+                                # predictive updates are best-effort and should not break the flow.
+                                logger.warning(
+                                    "Failed to decode JSON arguments for predictive tool '%s' (tool_call_id=%s).",
+                                    tool_name,
+                                    tool_call_id,
+                                )
+
+                        # Emit confirm_changes tool call
+                        confirm_id = generate_event_id()
+                        yield ToolCallStartEvent(
+                            tool_call_id=confirm_id,
+                            tool_call_name="confirm_changes",
+                            parent_message_id=flow.message_id,
+                        )
+                        confirm_args = {
+                            "function_name": tool_name,
+                            "function_call_id": tool_call_id,
+                            "function_arguments": json.loads(tool_call.get("function", {}).get("arguments", "{}")),
+                            "steps": [{"description": f"Execute {tool_name}", "status": "enabled"}],
+                        }
+                        yield ToolCallArgsEvent(tool_call_id=confirm_id, delta=json.dumps(confirm_args))
+                        yield ToolCallEndEvent(tool_call_id=confirm_id)
+                        flow.waiting_for_approval = True
+
+    # Close any open message
+    if flow.message_id:
+        yield TextMessageEndEvent(message_id=flow.message_id)
+
+    # Emit MessagesSnapshotEvent if we have tool calls or results
+    # Feature #5: Suppress intermediate snapshots for predictive tools without confirmation
+    should_emit_snapshot = flow.pending_tool_calls or flow.tool_results or flow.accumulated_text
+    if should_emit_snapshot:
+        # Check if we should suppress for predictive tool
+        last_tool_name = None
+        if flow.tool_results:
+            last_result = flow.tool_results[-1]
+            last_call_id = last_result.get("toolCallId")
+            last_tool_name = flow.get_tool_name(last_call_id)
+        if not _should_suppress_intermediate_snapshot(
+            last_tool_name, config.predict_state_config, config.require_confirmation
+        ):
+            yield _build_messages_snapshot(flow, snapshot_messages)
+
+    # Always emit RunFinished - confirm_changes tool call is complete (Start -> Args -> End)
+    # The UI will show confirmation dialog and send a new request when user responds
+    yield RunFinishedEvent(run_id=run_id, thread_id=thread_id)
@@ -6,21 +6,13 @@ import sys
 from typing import Any, TypedDict

 from agent_framework import ChatOptions
+from pydantic import BaseModel, Field

 if sys.version_info >= (3, 13):
    from typing import TypeVar
 else:
    from typing_extensions import TypeVar

-__all__ = [
-    "AGUIChatOptions",
-    "AgentState",
-    "PredictStateConfig",
-    "RunMetadata",
-]
-
-from pydantic import BaseModel, Field
-

 class PredictStateConfig(TypedDict):
    """Configuration for predictive state updates."""
@@ -63,6 +55,22 @@ class AGUIRequest(BaseModel):
        None,
        description="Optional shared state for agentic generative UI",
    )
+    tools: list[dict[str, Any]] | None = Field(
+        None,
+        description="Client-side tools to advertise to the LLM",
+    )
+    context: list[dict[str, Any]] | None = Field(
+        None,
+        description="List of context objects provided to the agent",
+    )
+    forwarded_props: dict[str, Any] | None = Field(
+        None,
+        description="Additional properties forwarded to the agent",
+    )
+    parent_run_id: str | None = Field(
+        None,
+        description="ID of the run that spawned this run",
+    )


 # region AG-UI Chat Options TypedDict
@@ -141,11 +141,14 @@ def make_json_safe(obj: Any) -> Any:  # noqa: ANN401
    if isinstance(obj, (datetime, date)):
        return obj.isoformat()
    if is_dataclass(obj):
-        return asdict(obj)  # type: ignore[arg-type]
+        # asdict may return nested non-dataclass objects, so recursively make them safe
+        return make_json_safe(asdict(obj))  # type: ignore[arg-type]
    if hasattr(obj, "model_dump"):
-        return obj.model_dump()  # type: ignore[no-any-return]
+        return make_json_safe(obj.model_dump())  # type: ignore[no-any-return]
+    if hasattr(obj, "to_dict"):
+        return make_json_safe(obj.to_dict())  # type: ignore[no-any-return]
    if hasattr(obj, "dict"):
-        return obj.dict()  # type: ignore[no-any-return]
+        return make_json_safe(obj.dict())  # type: ignore[no-any-return]
    if hasattr(obj, "__dict__"):
        return {key: make_json_safe(value) for key, value in vars(obj).items()}  # type: ignore[misc]
    if isinstance(obj, (list, tuple)):