## Why
Once multi-agent mode can be selected per turn, clients also need to
choose the initial selection when creating a thread and observe that
selection through lifecycle and settings APIs.
The selected value is intentionally distinct from the effective
model-visible value: no client selection is represented as `null`, even
though an eligible multi-agent v2 turn derives `explicitRequestOnly` as
its effective default.
## What changed
- Add the optional experimental `thread/start.multiAgentMode` parameter
and pass it through thread creation.
- Preserve an omitted initial value as an unset selection rather than
eagerly storing `explicitRequestOnly`.
- Apply an explicit `thread/start` selection to the first turn through
the session configuration established at thread creation.
- Restore the latest persisted effective mode as the selected baseline
on cold resume when rollout history contains one.
- Inherit the optional selected mode from a loaded parent when creating
related runtime threads.
- Return the current selected `multiAgentMode` from `thread/start`,
`thread/resume`, `thread/fork`, and thread settings, using `null` when
no mode is selected.
- Keep lifecycle reporting independent from model capability and feature
eligibility; core turn construction remains responsible for calculating
and persisting the effective mode.
## Not covered
- Clearing an existing loaded-session selection back to unset through
`turn/start`; omitted or `null` currently retains the session's
selection.
- A TUI control, slash command, or `config.toml` preference.
## Verification
- `CARGO_INCREMENTAL=0 just test -p codex-app-server-protocol`
- `CARGO_INCREMENTAL=0 just test -p codex-app-server multi_agent_mode`
The focused app-server coverage verifies explicit `thread/start`
initialization, first-turn prompting, nullable reporting for an omitted
selection, and retention of selections that are not currently
runtime-eligible.
## Stack
Stacked on #28685. This PR contains only the thread initialization and
lifecycle/settings API layer.
## Why
Multi-agent v2 currently carries an explicit-request-only delegation
rule in its static usage hint. That provides a safe default, but it
prevents clients from selecting proactive delegation per turn without
changing static guidance or rewriting prior model context.
This change makes delegation mode a session selection that can be
updated through `turn/start`, while deriving the effective model-visible
mode separately for each turn. Eligible multi-agent v2 turns remain
explicit-request-only unless proactive mode is both selected and
enabled.
## What changed
- Add the experimental `turn/start.multiAgentMode` parameter with
`explicitRequestOnly` and `proactive` values. Omission retains the
loaded session's current optional selection.
- Add the default-off `features.multi_agent_mode` feature gate. Eligible
multi-agent v2 turns use the selected mode when enabled; an unset
selection or disabled gate resolves to `explicitRequestOnly`.
- Treat mode prompting as inapplicable for multi-agent v1 and other
unsupported session configurations, producing no multi-agent mode
developer message rather than rejecting the turn.
- Move the explicit-request-only rule out of the static v2 usage hint
and into a bounded, tagged developer context fragment.
- Emit the effective mode in initial context and only when that
effective mode changes on later turns.
- Persist the effective mode in `TurnContextItem` as the durable
baseline for resume and context-update comparisons.
Historical rollout items are not rewritten. Later mode developer
messages establish the current rule incrementally.
## Not covered
- Initial selection through `thread/start` and selected-mode reporting
from thread lifecycle/settings APIs; those are isolated in the stacked
#28792.
- A TUI control or slash command for selecting the mode.
- Persisting a preferred mode to `config.toml`; selection remains
session/turn scoped.
- Changes to multi-agent concurrency limits, tool availability, or model
catalog capability declarations.
- Rewriting historical rollout prompt items. Cold resume restores the
latest persisted effective mode when available while leaving historical
developer messages intact.
## Verification
- `CARGO_INCREMENTAL=0 just test -p codex-core multi_agent_mode`
- Focused app-server coverage verifies that `turn/start.multiAgentMode`
produces proactive developer instructions for an eligible v2 turn.
## Stack
Followed by #28792, which adds `thread/start` initialization and
lifecycle/settings observability.
## Why
Remote environments registered through `environment/add` currently use
the fixed 10-second WebSocket connection timeout. Slow-starting
executors need a caller-selected connection window, but this should not
add retry policy or couple exec-server behavior to Core’s
`deferred_executor` feature.
Make the timeout an optional part of the existing experimental request.
Existing clients continue using the current default, while callers that
know an executor may take longer can request a larger window explicitly.
Depends on #28683.
## What changed
- Add optional `connectTimeoutMs` to `EnvironmentAddParams` and document
it in the app-server README.
- Pass the optional timeout through `EnvironmentRequestProcessor` into
one `EnvironmentManager::upsert_environment()` path; the manager applies
the existing default when it is omitted.
- Preserve the existing single-attempt lifecycle. The configured value
controls WebSocket connection and handshake time for both initial
connection and later reconnects; initialization retains its separate
timeout.
- Add an app-server integration test that sends the real JSON-RPC
request and verifies a stalled handshake observes the requested timeout.
## Test plan
- `just test -p codex-app-server-protocol`
- `just test -p codex-exec-server`
- `just test -p codex-app-server
environment_add_applies_connect_timeout`
## Rollout
This is additive and does not enable `deferred_executor`. Callers should
send a non-default timeout only after a compatible app-server is
deployed; omitted or `null` values retain the existing 10-second
default.
## Why
Remote environments may still be resolving when Codex creates a session
or turn. Waiting for the existing all-or-nothing environment snapshot
can hold startup until the selected environment is usable.
Behind the default-off `deferred_executor` feature, let callers take a
useful snapshot immediately: completed environments remain available
normally, while unfinished environments are reported without blocking
startup. With the feature disabled, snapshots preserve the existing
blocking behavior.
Depends on #28674.
## What changed
- Store one ordered list of selected environments in
`ThreadEnvironments`. Each selection owns one shared resolution that
produces its complete `TurnEnvironment`.
- Start new resolutions in the background with `remote_handle()`,
allowing snapshots and the future wait tool to share the same result
while cancellation follows the retained handles.
- Make `snapshot()` a read-only operation: nonblocking snapshots collect
completed resolutions and retain handles for unfinished ones, while
blocking snapshots await every resolution.
- Replace completed failed resolutions from the current manager entry
and log when failed environments are omitted.
- Return attached and starting environments as a point-in-time view, and
count starting environments when deciding whether a snapshot is
local-only.
- Keep existing consumers attached-only. `to_selections()` derives from
attached environments, so child threads do not inherit an environment
that is still starting.
## Test plan
- `just test -p codex-core environment_selection`
- `just test -p codex-core
deferred_executor_reaches_model_before_remote_environment_is_ready`
## Landing note
Keep `deferred_executor` disabled for slow-starting executors until
configurable `environment/add` connection timeouts and caller support
land. When enabled, an environment that attaches after session startup
may remain absent from environment-derived model context, tools,
instructions, skills, and related state until follow-up refresh work
lands.
## Why
Remote environments can be registered before their exec-server is first
used. Starting the connection at registration time uses that startup
window, while sharing one startup result prevents background work and
capability calls from opening competing connections.
Keep initial startup simple: each environment makes one connection
attempt using its configured transport timeout. A failed initial attempt
is final for that environment, while an environment that disconnects
after connecting can still recover on a later operation.
## What changed
- Start URL and Noise environments in the background when they are added
to `EnvironmentManager`. Provider snapshots are fully validated before
connection work begins.
- Share one initial connection attempt and its saved result across
metadata, process, filesystem, and HTTP callers.
- Keep configured stdio environments lazy until first use so
registration does not launch a process.
- Tie background startup work to the environment lifetime so replacing
or dropping an environment cancels unfinished work.
- After an established client disconnects, share one fresh connection
attempt across concurrent callers. A failed attempt fails the current
operation without permanently preventing a later attempt.
- Store the shared lazy client directly on `Environment` and expose
small methods for starting, observing, and awaiting startup.
## Test plan
- `just test -p codex-exec-server`
- `just test -p codex-app-server
turn_start_resolves_sticky_thread_local_environment_and_turn_overrides`
## Why
Plugin-install preflight and the actual OAuth login flow used different
discovery implementations. Preflight had a Codex-specific implementation
that only queried authorization-server metadata on the MCP host, while
login already used the upstream `rmcp` Rust MCP SDK. As a result,
servers that advertise a separate authorization server through RFC 9728
Protected Resource Metadata were classified as OAuth-unsupported during
plugin installation, so login was skipped.
## What changed
- delegate plugin-install OAuth discovery to
`rmcp::transport::AuthorizationManager`, the same implementation used by
the login flow
- let `rmcp` follow Protected Resource Metadata first and perform direct
RFC 8414 authorization-server discovery when protected-resource
discovery does not yield usable metadata
- retain Codex's existing HTTP headers, timeout, `no_proxy` behavior,
and scope normalization around that discovery
- add unit coverage and a pure-MCP plugin-install integration test that
proves the protected-resource path reaches OAuth client registration
This only changes shared MCP OAuth discovery. App declarations and
`appsNeedingAuth` behavior are unchanged.
## Verification
- `just test -p codex-rmcp-client auth_status`
- `just test -p codex-app-server plugin_install_starts_mcp_oauth`
- real plugin-install smoke test with an isolated `CODEX_HOME`: both
DigitalOcean MCP servers started OAuth callback listeners, while Linear
continued to start its existing direct-discovery OAuth flow
## Why
Remote v2 compaction can return replacement-history items without IDs.
Because replacement history is installed directly, those items bypass
normal history preparation and remain ID-less in later Responses
requests even when the `item_ids` feature is enabled.
## What changed
- Pass the active `TurnContext` into `replace_compacted_history`.
- When `item_ids` is enabled, assign missing IDs before installing and
persisting replacement history.
- Rebuild `CompactedItem` from the prepared history so live and
persisted replacement histories match.
- Add integration coverage requiring IDs on every ID-capable input item
in the initial, remote v2 compaction, and post-compaction requests.
## Test plan
- `just test -p codex-core response_item_ids`
- `just test -p codex-core websocket_v2_test_codex_shell_chain`
- `just test -p codex-core remote_compaction_parity_pre_turn_auto`
- `just test -p codex-app-server
thread_inject_items_adds_raw_response_items_to_thread_history`
## Summary
- expose `clock.curr_time` when current-time reminders are enabled
- query the session's configured time provider with the calling thread
id
- return the existing UTC reminder text for direct model calls
- return `{ "current_time": "YYYY-MM-DD HH:MM:SS UTC" }` in Code Mode
Clock lookup failures remain fatal, matching pre-inference reminder
behavior.
## Testing
- `just test -p codex-core current_time_tool_returns_the_latest_time`
- `just test -p codex-core
code_mode_current_time_returns_structured_result`
- `just fix -p codex-core`
## Summary
- skip the legacy `openai-curated` startup repository sync when remote
plugins are enabled and the current auth uses the Codex backend
- keep the curated sync for API-key, Bedrock, and unauthenticated
sessions that fall back to the local marketplace
- preserve configured marketplace upgrades and all remote plugin startup
warmups
## Why
The remote catalog owns plugin discovery and materialization only when
it is usable for the current auth mode. Starting the legacy curated
repository sync in that case performs an unnecessary Git/HTTP/archive
download and cache refresh. API-key and Bedrock sessions still require
the local curated marketplace, so they must continue syncing it.
## User impact
Codex startup no longer downloads or refreshes the local
`openai-curated` snapshot when the remote catalog is active. Behavior is
unchanged for auth modes that use the local curated marketplace.
## Validation
- `just fmt`
- `git diff --check`
Rust tests were not run per the repository's local verification policy
for this narrow conditional change.
## Why
Client-created response items enter history without IDs, so their
identity is lost across rollout persistence and resume. IDs should be
assigned once at the history-recording boundary, while IDs returned by
the server must remain unchanged.
The Responses API validates item IDs using type-specific prefixes.
Locally generated IDs therefore use the matching prefix plus a
hyphenated UUIDv7, keeping them valid while distinguishable from
server-generated IDs. Because this changes persisted history and
provider request shapes, the behavior is opt-in behind the
under-development `item_ids` feature. Compaction triggers remain request
controls whose API shape does not accept an ID.
## What changed
- Register the disabled-by-default `item_ids` feature and expose it in
`config.schema.json`.
- Make supported optional `ResponseItem` IDs serializable and expose
them in the generated app-server schemas.
- When `item_ids` is enabled, assign an ID during conversation-history
preparation if an item has no ID.
- Generate type-prefixed, hyphenated UUIDv7 IDs using the Responses API
item conventions.
- Preserve existing server IDs without rewriting them.
- Persist assigned IDs in rollouts and include them in subsequent
Responses requests.
- Remove the unsupported ID field from `CompactionTrigger` and document
why it has no ID.
- Add integration coverage for enabled ID persistence, preservation of
server IDs, and omission of generated IDs while the feature is disabled.
`prepare_conversation_items_for_history` is the single response-item ID
allocation boundary.
## Test plan
- `just test -p codex-features`
- `just test -p codex-core
response_item_ids_persist_across_resume_and_preserve_server_ids`
- `just test -p codex-core
non_openai_responses_requests_omit_item_turn_metadata`
- `just test -p codex-core
resize_all_images_prepares_failures_before_history_insertion`
- `just test -p codex-protocol`
- `just test -p codex-app-server-protocol`
- `just test -p codex-api azure_default_store_attaches_ids_and_headers`
## Summary
- Remove the realtime `architecture` selector from core protocol,
app-server protocol, config parsing, generated schemas, and callers.
- Always create WebRTC realtime calls with the AVAS query params:
`intent=quicksilver&architecture=avas`.
- Keep direct websocket realtime behavior on the existing config/default
path, while WebRTC starts without an explicit version now default to
realtime v1 because AVAS requires v1.
## Notes
- WebRTC realtime now means AVAS. If a caller explicitly asks to start
WebRTC with realtime v2, Codex rejects that request because the AVAS
WebRTC path only supports realtime v1. Websocket realtime is separate
and can still use realtime v2.
- The old `[realtime] architecture = "realtimeapi" | "avas"` config knob
is removed. Local configs that still set it will need to delete that
line.
- Some app-server tests that were only trying to exercise realtime v2
protocol behavior now use websocket transport, because WebRTC is
intentionally locked to AVAS/v1. Separate WebRTC tests cover the AVAS
query params, v1 startup, SDP flow, and sideband join.
## Validation
- Merged fresh `origin/main` at `83e6a786a2`.
- `just fmt`
- `just write-config-schema`
- `just write-app-server-schema`
- `git diff --check`
- `just test -p codex-api -p codex-core -p codex-app-server-protocol -p
codex-app-server realtime` (176 passed)
- `just test -p codex-protocol -p codex-config` (413 passed)
Summary
- Refresh the installed remote-plugin snapshot and Codex Apps tools
after completing a remote JIT install.
- Gate `completed: true` on every expected `app_connector_id` appearing
after the uncached `tools/list` refresh, while continuing to skip local
bundle verification for server-side installs.
- Keep the cached recommendations response and filter refreshed
installed remote IDs locally, so this does not add another
recommendations fetch.
- Add regression coverage for tools appearing after the hard refresh and
remaining absent after the refresh. The resumed model request sees the
refreshed tool router when installation completes.
Root Cause
- Remote suggestions from `openai-curated-remote` returned `true` before
taking the existing connector refresh path, leaving the resumed turn
with the pre-install Apps tool catalog.
Validation
- `just test -p codex-core request_plugin_install`
- `just test -p codex-core-plugins
recommended_plugin_candidates_filter_installed_and_disabled_plugins`
- `just test -p codex-core-plugins`
- `just fix -p codex-core-plugins`
- `just fix -p codex-core`
- `just fmt`
- `just test -p codex-core` was not fully clean locally: 2,729 passed,
26 failed, and 16 skipped. The failures were dominated by local
Seatbelt/network/timing issues, including plugin-install timeouts under
full-suite contention; the focused plugin-install runs pass.
## Why
The token-budget context currently identifies a context window by its
thread-local sequence number. A UUIDv7 gives the model a stable opaque
identity that remains fixed for a window and rotates when compaction or
`new_context` starts the next one.
## What changed
- Preserve the existing monotonic value as `window_number` and add a
UUIDv7 `window_id` to `CompactedItem`.
- Generate and rotate the UUID with auto-compaction window state,
persist it alongside the number, and reconstruct it on resume and
rollback.
- Accept legacy compacted rollout records where the numeric `window_id`
represented the window number.
- Use the UUID only in token-budget context; existing request headers
and metadata continue using `thread_id:window_number`.
## Testing
- `just test -p codex-protocol compacted_item::tests`
- `just test -p codex-core token_budget`
## Summary
- Preserve raw plugin skill-root snapshots in the matching loaded-plugin
cache entry, keyed by the effective plugin root identity including
namespace.
- Pass those snapshots through `SkillsLoadInput` as an optional preload,
so session startup reuses plugin parsing while ordinary skill loads pass
`None`.
- Keep plugin skill loading cohesive: the existing loaders accept the
optional snapshots directly, and uncached or marketplace-detail paths do
not create a cache.
## Why
Plugin discovery already parses plugin skills to determine available
capabilities. Cold session startup then scanned and parsed the same
roots again while building the skills snapshot.
This solves the same duplicate-work problem as #28623 while keeping
ownership narrow: `PluginsManager` creates and owns
`PluginSkillSnapshots` only for its loaded-plugin cache entry;
`SkillsService` consumes an optional clone. Entry replacement or
clearing naturally drops the snapshots, with no separate generation,
capacity policy, or watcher coupling.
## Validation
- `cargo clippy -p codex-core-skills --all-targets -- -D warnings`
- `just test -p codex-core-plugins
skills_service_reuses_skills_parsed_during_plugin_load`
- `just test -p codex-core-skills
namespaces_plugin_skills_using_provided_namespace`
- `just fmt`
## Why
We need to avoid resolving shells on the app-server's host for remote
environments. We might make it possible to do fancier shell resolution
from remote envs but for now just require the model to produce a shell
that matches the environment's default.
This gets my e2e demo working for shell commands after #28854 moved
shell resolution to PathUri and caused remote envs to hit the fallback
shell when the shell wasn't available on the host.
## What
Remote `exec_command` calls now accept only the environment's reported
default shell name or exact path, and execute with that reported path.
Other explicit shells return a concise error. A Wine-backed integration
test covers explicit PowerShell execution in the Windows cwd.
## Why
No need to do path contortions when it's for our own logs.
## What
Follow up on a previous PR's nit and update the path-types skill for
future reference.
## Why
`child_agents_md` is a disabled, under-development experiment that adds
a second model-visible explanation of hierarchical `AGENTS.md` behavior.
Keeping it leaves unused prompt, configuration, documentation, and test
surface.
## What changed
- remove the `ChildAgentsMd` feature and `child_agents_md` config schema
entry
- remove the hierarchical prompt asset, export, and instruction
injection
- remove feature-specific tests and documentation
- keep the generic unstable-feature warning coverage using
`apply_patch_streaming_events`
Normal project `AGENTS.md` discovery and composition are unchanged.
## Testing
- `just test -p codex-features`
- `just test -p codex-prompts`
- `just test -p codex-core agents_md`
- `just test -p codex-core unstable_features_warning`
## Summary
Support marketplace plugins whose source directory does not include a
discoverable plugin manifest. Metadata-rich `marketplace.json` entries
now act as fallback plugin manifests for listing, local detail reads,
install, and non-curated cache refresh.
The fallback preserves marketplace-entry plugin fields wholesale, then
adds the small Codex-facing compatibility bridge for presentation
metadata. A real source `plugin.json` always wins when present.
## Details
- Capture flattened marketplace-entry fields into
`MarketplacePluginManifestFallback`, preserving fields such as
`version`, `description`, `skills`, `mcpServers`, `apps`, `hooks`,
`agents`, `commands`, `strict`, `author`, and future manifest fields
without a per-field translation list.
- Bridge Claude-style top-level `displayName`, `author.name`,
`homepage`, and marketplace `category` into Codex's nested `interface`
fields only when the nested values are absent.
- Treat fallback metadata as installable only when the marketplace entry
contributes metadata beyond bare `name` and `source`; existing
missing-manifest behavior remains for metadata-free entries.
- Read local plugin details from the already parsed fallback manifest,
including fallback-declared app and MCP paths, instead of rereading only
an on-disk manifest.
- Pass fallback contents into `PluginStore`, which validates them and
injects `.codex-plugin/plugin.json` into Store's existing atomic copy.
Local marketplace source directories are never mutated, and the fallback
path no longer needs an additional staging directory.
- Keep Git source materialization unchanged; Git clones still use the
existing marketplace source staging area before Store installation.
## Why
Make it possible to load AGENTS.md from remote exec-servers whose OS is
different than app-server.
## What
- keep `AGENTS.md` discovery and provenance as `PathUri`, with
root-aware parent and ancestor traversal
- expose lifecycle instruction sources as legacy app-server path strings
in events while retaining `PathUri` internally
- preserve and test mixed POSIX and Windows paths in model context and
TUI status output
- cover remote Windows loading end to end by seeding the Wine prefix
through host filesystem APIs
- fix bug in `PathUri`'s parent() implementation that would erase
Windows drive letters
## Summary
- preserve the original HTTP status when a remote plugin bundle download
returns a non-success response
- retain at most 8 KiB of the error response body and annotate
truncation or body-read failures
- add regression coverage for an oversized error response
## Root cause
The non-success response path reused the normal size-limited body
reader. When an error response exceeded 8 KiB, that reader returned
`DownloadTooLarge` before the code constructed `DownloadStatus`, masking
the upstream HTTP status and response context.
## Impact
Remote plugin installation failures now retain the actionable upstream
HTTP status without allowing unbounded error bodies into logs.
## Validation
- `just test -p codex-app-server
plugin_install_preserves_status_when_remote_bundle_error_body_is_too_large`
- `just fmt`
- `git diff --check`
Summary
- Stop treating Codex Apps MCP tools with
`_meta._codex_apps.synthetic_link: true` as evidence that a connector is
accessible in `app/list`.
- Preserve synthetic tools in the agent-facing MCP connector set so they
remain available for install/auth flows.
- Keep the app-list accessibility cache limited to connectors backed by
at least one non-synthetic tool.
- Add focused regression coverage for both sides of the boundary.
Validation
- `just fmt`
- `just test -p codex-core
synthetic_links_are_exposed_to_the_agent_but_not_accessible_in_app_list`
- `git diff --check`
- A crate-wide `just test -p codex-core` run completed with 2,699
passing and 51 unrelated local sandbox/state failures, primarily state
DB migration races (`UNIQUE constraint failed:
_sqlx_migrations.version`).
## Stack
This is PR 2 of the simplified HAI single-run-task stack:
- [#19047](https://github.com/openai/codex/pull/19047) Agent Identity
assertion and task-registration primitives, including the shared
run-task helper used by existing Agent Identity JWT auth.
- [#19049](https://github.com/openai/codex/pull/19049)
Disabled-by-default ChatGPT auth opt-in that provisions/reuses persisted
Agent Identity runtime auth and its single run task.
- [#19051](https://github.com/openai/codex/pull/19051) Run-scoped
provider auth that uses one backend-owned task id for first-party
inference and compaction requests.
[#19054](https://github.com/openai/codex/pull/19054) collapsed out of
the active stack because the simplified design no longer needs a
separate background/control-plane task helper.
## Summary
This PR adds the disabled-by-default path for normal ChatGPT-login Codex
sessions to obtain Agent Identity runtime auth through the Codex
backend. Existing Agent Identity JWT startup mode remains a separate
path and does not require the feature flag.
What changed:
- adds the experimental `use_agent_identity` feature flag and config
schema entry
- adds an explicit `AgentIdentityAuthPolicy` so call sites choose
`JwtOnly` or `ChatGptAuth` instead of passing a bare boolean
- stores standalone Agent Identity JWT credentials separately from
backend-registered Agent Identity records
- persists the registered Agent Identity record, private key, and single
run task id in `auth.json` so process restarts reuse the same identity
- derives the agent/task registration base URL from ChatGPT/Codex auth
config while keeping JWT JWKS lookup separate
- provisions and caches ChatGPT-derived Agent Identity runtime auth when
`use_agent_identity` is enabled
- reuses the shared run-task registration helper from PR1 rather than
adding a second task-registration path
This PR intentionally does not switch model inference over to
`AgentAssertion` auth. The provider-auth integration lands in the next
PR.
## Testing
- `just test -p codex-login`
## Summary
[@milanglacier reported in
#28653](https://github.com/openai/codex/issues/28653) that the active
mention candidate is hard to distinguish. I suspect [@binbjz’s #28500
report](https://github.com/openai/codex/issues/28500) _(where arrow-key
navigation appeared not to work)_ may describe the same presentation
problem: the selection may have been changing, but the UI was not
showing the active row clearly in their terminal. This PR makes two
small changes to the selection indication behavior:
- Reserve a two-character gutter and mark the active candidate with `> `
for color-agnostic indicator coverage.
- Apply the shared theme-aware accent to the entire selected row for
extra emphasis.
- Update the existing popup snapshot.
Reverse-video styling was considered, but avoided it because it is
overly dependent on the user’s terminal palette.
<img width="2046" height="482" alt="image"
src="https://github.com/user-attachments/assets/b5eb62c3-fd24-4c09-906e-7bd66913b5c6"
/>
## Testing
- `just test -p codex-tui default_unified_mention_popup_snapshot`
- `just clippy -p codex-tui`
- `just fmt`
- Compiled `codex-cli` and tested the unified mentions picker in the
terminal.
## Summary
- remove the duplicated originator-specific connector ID denylists
- stop filtering connector directory/accessibility results and
live/cached Codex Apps MCP tools by hardcoded connector ID
- remove the now-unused `codex-login` dependency from
`codex-utils-plugins`
- update regression coverage so formerly blocked connector IDs are
preserved
## Why
The client-side policy was duplicated across crates, used opaque IDs
without ownership or expiry information, and could drift between app
listing and MCP tool behavior. Server-provided visibility,
authorization, plugin discoverability, accessibility, enabled-state
handling, and consequential-tool approval templates remain unchanged.
## Validation
- `just fmt`
- `just bazel-lock-update`
- `just bazel-lock-check`
- `git diff --check`
- confirmed the final diff contains no hardcoded denylist symbols
A targeted `codex-mcp` test build spent an unusually long time in local
compilation/linking. Its first attempt exposed a test-only `PartialEq`
assertion issue, which was corrected. A follow-up non-linking `cargo
check -p codex-mcp --tests` was still running when this draft was
opened; CI should provide the complete Rust validation.
## Why
`on-request` approval policy text is currently tuned for user-reviewed
approvals. For auto-reviewed productivity runs, likely sandbox blocks
should be escalated earlier so commands that need remote services,
authentication, or other out-of-sandbox access do not first fail or hang
inside the sandbox.
## What changed
- Adds a separate `on_request_auto_review.md` permissions prompt
selected for `AskForApproval::OnRequest` with
`ApprovalsReviewer::AutoReview`.
- Keeps the normal user-reviewed `on-request` wording unchanged.
- Makes the `When to request escalation` bullets more explicit about
likely sandbox blocks, network access, remote
auth/cluster/cloud/database access, out-of-sandbox environment access,
git operations that may write lock files, and short-timeout reruns after
likely sandbox-blocked attempts.
- Omits approved command prefix and `prefix_rule` guidance for the
auto-review on-request prompt.
- Adds prompt tests covering the auto-review path, normal on-request
wording, and inline permission request behavior.
## What
Server should request:
```
{
"id": 42,
"method": "currentTime/read",
"params": {
"threadId": "11111111-1111-1111-1111-aaaaafdc2c11"
}
}
```
Client should respond with something like:
```rust
{
"id": 42,
"result": {
"currentTimeAt": 1781717655
}
}
```
## Why
Sessions configured with `clock_source = "external"` need a
thread-specific external time source before inference. The system clock
remains the default production provider.
## Validation
- `cargo test -p codex-app-server-protocol`
- `cargo test -p codex-app-server --test all
current_time_read_round_trip_adds_reminder_to_model_input`
- `cargo test -p codex-app-server
first_attestation_capable_connection_for_thread_only_uses_thread_subscribers`
- `cargo test -p codex-analytics`
- `just fix -p codex-app-server-protocol`
- `just fix -p codex-app-server`
Stacked on #28824.
## Why
Allows the model to edit files that are hosted on a different OS than
where app-server is running.
## What
* Use `PathUri` for apply_patch-internal data structures
* Limit `PathUri` -> `AbsolutePathBuf` conversion to cases where the
inferred path convention matches the host OS, allows requiring valid
paths to pass to perms check
* Adds `PathConvention::path_segments()` for iterating over path
segments regardless of OS
* Handle cross-platform relative paths in path filename parsing for
sniffing a shell
* Ensure we can apply patches in the wine e2e test
## Why
`built_tools` runs for every sampling request, and local plugin
discovery was repeatedly rereading plugin manifests, skills, MCP
configuration, and app declarations to build the same tool-suggest
metadata.
That source-derived metadata is stable until the existing plugin manager
reloads its cache. Runtime eligibility still needs to reflect the
current install, disable, policy, app-overlap, and authentication state.
## What changed
- Add a bounded, in-memory tool-suggest metadata cache owned by
`PluginsManager`.
- Key cached metadata by plugin identity and source, while applying
authentication routing each time the metadata is projected.
- Invalidate the metadata alongside the existing loaded-plugin cache,
including its normal configuration, marketplace refresh, and
remote-installed-plugin invalidation paths.
- Guard against an in-flight load repopulating stale metadata after
invalidation.
- Keep marketplace membership and all runtime eligibility filtering live
rather than introducing a separate catalog or revision model.
## Impact
Repeated sampling requests reuse already-loaded plugin capability
metadata while retaining the existing plugin-manager lifecycle as the
single freshness boundary.
## Validation
- `just test -p codex-core-plugins` — 252 passed
- Added focused coverage for cache invalidation and authentication
reprojection.
Stacked on #28822.
## Summary
- add a host-injectable current-time provider with a built-in system
implementation
- record UTC developer reminders in history immediately before due model
requests
- keep cadence state per session and force a refresh after compaction
This does NOT include the app server client <-> server clock logic. This
PR is only for the reminder message & system clock that will be used in
prod.
## Testing
- `just test -p codex-core varlatency_`
- `just clippy -p codex-core -p codex-app-server -p codex-mcp-server -p
codex-thread-manager-sample`
- `just fmt`
Make `ListItemsParams::turn_id` optional so callers can list persisted
items across an entire thread or narrow the result to one turn. This
aligns the thread-store API and documentation with thread-wide item
listing while preserving the optional turn-filter behavior for
implementations.
## Stack
Depends on #28746. This PR implements shared rollout-budget accounting
and model-visible reminders using the configuration defined in #28746.
# Description / Main changes to Core:
`AgentControl` will now be the area where "rollout level" features &
accounting will have to live. It is incorrectly named for this
responsibility, but I think it can hold all the necessary shared state &
features (rollout token budget, mutliple thread interruption
responsibilitym etc)
In this PR, we have one "token ledger" that each thread will subtract
from when sampling. The "charge" will occur when response.completed() is
done and the calculation will be done on the responses api usage
carrier. The calculation will weigh sampling and pre-fill tokens as
specified.
Every time the budget crosses the configured reminder threshold, a
developer message is appended before the thread's next request
This remaining budget will _always_ be restated/reminded after a
compaction event.
Expiration and fan-out interruption will be in the stacked follow-up
(and also live in Agent Control).
## Reminders
"You have weighted {session_tokens_left} tokens left in the shared
session token budget."
The first request in each thread context receives the current remainder.
Later reminders are emitted after aggregate weighted usage crosses a
configured interval. If several intervals are crossed before a thread
sends another request, Core inserts one reminder with the latest
remainder.
Compaction response usage is charged before the next context starts. The
next reminder is appended after the compaction summary, leaving the
initial context content stable.
## Tests
Integration coverage verifies:
- weighted output and non-cached input accounting
- initial and periodic reminders
- shared accounting between a root and sub-agent
- post-compaction remainder and message placement
Local checks:
- `just fmt`
- `just test -p codex-core rollout_budget`
- `git diff --check`
The full workspace test suite was not run locally.
## What
Deliver the scripted realtime notification batch after the assistant
text append request instead of after the preceding developer text append
request.
## Why
The batch ends with an upstream error that closes the realtime
conversation. When it is emitted after the developer append, it races
the subsequent assistant append: the app-server RPC can acknowledge the
append before its downstream WebSocket send completes, and the test
intermittently observes three requests instead of four.
Making the fake server wait for the assistant append before emitting the
terminal batch establishes the ordering the test asserts without sleeps
or production-code changes.
## Validation
- `git diff --check`
- CI (the failure is timing-dependent and most reproducible in the
Windows Bazel shard)
## Why
Codex Desktop repairs sandbox-user read/execute access for binaries
copied to `%LOCALAPPDATA%\OpenAI\Codex\bin`, but Computer Use launches
its bundled Node runtime from `%LOCALAPPDATA%\OpenAI\Codex\runtimes`.
On fresh Windows installations, `CodexSandboxUsers` may therefore be
unable to execute the bundled Node binary. The command runner starts,
but `CreateProcessAsUserW` fails with error 5 (`ACCESS_DENIED`), causing
the Node REPL to exit before Computer Use can discover applications.
This is a follow-up to #21564, which added the original runtime `bin`
ACL repair.
## What changed
- Expand the Codex Desktop runtime ACL roots from only `bin` to both
`bin` and `runtimes`.
- Apply the existing inherited read/execute ACL repair to each runtime
directory when it exists.
- Rename the setup helper to reflect that it now handles multiple
runtime paths.
## Validation
- `cargo fmt -- --check`
- `just test -p codex-windows-sandbox` was run: 113 tests passed and
five environment-dependent legacy execution tests failed because
`CreateRestrictedToken` returned error 87.
## Summary
- Initialize stderr tracing and the configured OpenTelemetry provider
for local and remote `codex exec-server` startup.
- Instrument the local and remote server entrypoints with a root runtime
span.
- Keep raw Noise environment, registration, and stream identifiers out
of exported spans while preserving them in local debug events.
- Keep telemetry setup in a focused CLI module instead of growing the
top-level command entrypoint.
## Stack
- Previous: none (`#27058` has merged)
- Next: #27466
## Validation
- `just test -p codex-exec-server --lib` (139 passed)
- `just test -p codex-cli --test exec_server` (3 passed)
- `just bazel-lock-check`
- `just fix -p codex-exec-server -p codex-cli`
- `just fmt`
---------
Co-authored-by: Richard Lee <richardlee@openai.com>
Fixes#28263.
## Why
When a thread starts with `/goal`, the goal extension can update SQLite
goal state before the thread has any user-turn rollout items.
`thread/list` and `thread/search` rely on persisted listing metadata, so
a goal-first live thread could be absent from app-server listings after
restart even though the goal itself existed.
This regressed when goal handling moved out of core: the core path wrote
the goal update through the live thread rollout path, while the
extension-backed app-server path only updated goal state and emitted the
live notification.
## What
- Add `GoalSetOutcome::thread_goal_updated_item()` so the goal extension
owns the canonical `ThreadGoalUpdated` rollout item shape.
- Expose a narrow `CodexThread::append_rollout_items()` helper that
appends through the live thread and keeps derived SQLite metadata in
sync.
- When app-server sets a goal on an active live thread, persist the goal
update through that live-thread path.
- Add an app-server regression test that starts a live thread with
`thread/goal/set` and verifies it appears in state-DB-only
`thread/list`.
## Verification
- `env -u CODEX_SQLITE_HOME just test -p codex-app-server
goal_first_live_thread_appears_in_state_db_thread_list`
## Summary
- keep context injection on a single ContextContributor trait
- split context injection into thread-scoped and turn-scoped
contribution methods
- wire turn-scoped fragments into initial context assembly so extensions
can contribute context from turn-local state
Scope MCP sandbox metadata to the MCP server's owning environment.
Previously, `codex/sandbox-state-meta` always used the turn's primary
cwd and rebuilt a legacy sandbox policy from that cwd. That can be wrong
for MCP servers owned by a different execution environment.
This now sends the owning environment cwd as a `file:` URI in
`sandboxCwd`, keeps `permissionProfile` as the permission source of
truth, and omits sandbox-state metadata when a non-default server
environment is not selected for the turn. Local/default MCP servers keep
the existing fallback cwd behavior.
Tests:
- `just fmt`
- `just bazel-lock-update`
- `just bazel-lock-check`
- `just test -p codex-mcp`
- `just test -p codex-core mcp_sandbox_cwd`
- `cargo build -p codex-rmcp-client --bin test_stdio_server`
- `just test -p codex-core
stdio_mcp_tool_call_includes_sandbox_state_meta`
## Summary
Remote stdio MCP servers send tool calls by writing JSON-RPC bytes
through `process/write`.
When the exec-server websocket drops at the wrong time, the remote
process can survive session recovery, but the stdin write can still fail
back to RMCP as a transport send error. RMCP then closes the stdio MCP
transport, so tools like `node_repl` are lost even though the
process/session recovery path is working.
This changes `process/write` to be safe to retry across exec-server
recovery:
- adds a required `writeId` to `process/write`
- retries remote `Session::write` with the same `writeId` after
reconnect
- remembers accepted write ids per process so duplicate retries return
`Accepted` without writing the same bytes to child stdin again
- covers both the client retry path and server-side write id dedupe with
tests
In simple terms:
```text
before:
write to MCP stdin -> websocket closes -> write errors -> RMCP closes node_repl
after:
write to MCP stdin -> websocket closes -> reconnect -> retry same writeId
server either writes once or recognizes it already did
```
Fixes#28104.
## Summary
Active `/goal` turns should leave the persisted goal paused whenever the
TUI interrupts the running turn. The bug in #28104 showed this most
visibly through `Esc`: some interrupt paths aborted the turn without
updating the goal status, so the goal could remain active and continue
automatically.
This change makes `ChatWidget` pause an active goal before the TUI sends
an interrupt from the status-row path, the pending-steer path, `Ctrl+C`,
or a request-user-input overlay. The modal overlay now reports whether a
key will interrupt the turn, which keeps modal `Esc` and `Ctrl+C`
behavior aligned with the normal interrupt paths.
## Manual Testing
Built the local CLI with `just codex --help`, then launched the local
TUI with goals enabled. Started an active `/goal` turn and interrupted
it with `Esc`, then resumed and repeated with `Ctrl+C`; both paths
showed `Goal paused`, the interrupted-conversation message, and the
`Goal paused (/goal resume)` footer. I also stopped the background
terminal and exited the TUI cleanly after the run.
I did not find a reliable standalone manual path to force the
request-user-input overlay case, so that path is covered by the focused
automated test.
## Summary
This keeps the apply_patch approval tests focused on approval behavior
instead of macOS sandboxed filesystem helper startup.
The changed cases still force patch approval with `UnlessTrusted`, but
use `DangerFullAccess` after approval so the patch write is direct and
cheap. Workspace-write and sandbox-helper behavior remain covered by the
filesystem and apply_patch sandbox tests.
## Why
Prepare network approval scoping to distinguish execution environments
without changing behavior yet.
## What changed
- Add optional environment IDs to network policy requests.
- Add optional network environment IDs to exec and sandbox request
structs.
- Thread default None values through existing construction points.
- Fix stale constructor call sites that caused the CI compile failures.
## Not included
- Per-environment proxy listeners.
- Network approval cache or prompt behavior changes.
- Ambiguous request attribution handling.
Those behavior changes moved to stacked follow-up #28899.
## Validation
- just fmt
- CI will run tests and clippy
## What
This PR defines the structured configuration contract for shared rollout
token budgets (across ALL agent threads under 1 rollout).
```toml
[features.rollout_budget]
enabled = true
limit_tokens = 100000
reminder_interval_tokens = 10000
sampling_token_weight = 1.0
prefill_token_weight = 0.1
```
The reminder interval defaults to 10% of the rollout limit. Sampling and
prefill weights default to `1.0`.
## Scope
This PR only defines and validates configuration. It does not track
usage, inject reminders, or stop a rollout. Accounting and reminders are
implemented in the stacked follow-up #28494.
The existing `token_budget` feature remains unchanged. `rollout_budget`
has its own feature key and configuration type.
## Tests
The config test verifies that the structured fields resolve into
`RolloutBudgetConfig` and do not enable the existing `token_budget`
feature.
Local checks:
- `just write-config-schema`
- `just test -p codex-core load_config_resolves_rollout_budget`
- `cargo check -p codex-thread-manager-sample`
- `git diff --check`
The full workspace test suite was not run locally.
## What changed
- retain the parsed plugin manifest namespace on loaded plugins
- carry that namespace through `PluginSkillRoot` and `SkillRoot`
- use the provided namespace when qualifying plugin skill names
- include the namespace in the skills cache key
## Why
Plugin loading has already parsed `plugin.json`, but skill parsing
currently walks every `SKILL.md` ancestor and probes/reads the manifest
again to reconstruct the same namespace. Passing the parsed namespace
removes those repeated filesystem calls, which are particularly costly
on remote filesystems.
Context:
https://openai.slack.com/archives/C0ARA9GF5D4/p1781639496496439?thread_ts=1781202444.891669&cid=C0ARA9GF5D4
## Impact
Plugin skill names remain unchanged. A regression test uses a
deliberately different on-disk manifest name to verify that plugin roots
use the provided parsed namespace.
## Validation
- `just test -p codex-core-skills -p codex-core-plugins -p codex-plugin
-p codex-utils-plugins` (352 passed)
- `just fix -p codex-core-skills -p codex-core-plugins -p codex-plugin
-p codex-utils-plugins`
- `just fmt`
## What changed
- promote plugin config loading to an info-level `plugins_for_config`
span
- promote skill config loading to an info-level `skills_for_config` span
- attach stable OpenTelemetry names to both spans
## Why
`session_init.plugin_skill_warmup` currently combines plugin loading and
skill loading, which makes cold-start traces unable to identify which
phase dominates. These child spans preserve the existing aggregate while
making the two costs independently visible.
Context:
https://openai.slack.com/archives/C0ARA9GF5D4/p1781639496496439?thread_ts=1781202444.891669&cid=C0ARA9GF5D4
## Impact
This is observability-only. It does not change plugin or skill loading
behavior.
## Validation
- `just test -p codex-core-skills -p codex-core-plugins` (347 passed)
- `just fmt`
## Why
App-server must report command events containing foreign-platform paths
without changing existing client or rollout path-string formats.
## What changed
- retain `PathUri` through exec command begin/end events
- convert cwd values to `LegacyAppPathString` at the app-server
compatibility boundary
- drop command actions with foreign paths and log them
- serialize rollout-trace cwd values using their inferred native path
representation
- restore Wine coverage for retained Windows cwd values and successful
completion
## Summary
Allow plugin manifests to declare `skills` as either a single path
string or an array of path strings in the core plugin loader.
## Why
Some plugin packages need to expose skills from more than one directory.
Before this change, `plugin.json` only accepted a single string for
`skills`, so manifests like this were ignored as an invalid `skills`
shape:
```json
{
"skills": ["./skills/abc", "./skills/edk"]
}
```
This keeps the existing single-string form working while adding support
for the list form. The final scope is intentionally limited to the core
plugin manifest/load path for `skills`; `apps`, file-backed
`mcpServers`, and the bundled plugin-creator assets are unchanged in
this PR.
## What changed
- Parse `skills` as either a string or an array of strings in
`plugin.json`.
- Store resolved skill paths as a list in `PluginManifestPaths`.
- Load manifest-declared skill roots in addition to the default
`./skills` root.
- Deduplicate exact duplicate skill roots before loading.
- Rely on existing skill-loader dedupe by canonical `SKILL.md` path for
overlapping roots such as `./skills` plus `./skills/abc`.
- Update plugin manifest tests to cover:
- single string `skills`
- list of string `skills`
- duplicate skill roots
- `./skills` as a manifest path
- explicit child roots like `./skills/abc` and `./skills/edk`
- overlapping-root dedupe
## Validation
- `just test -p codex-plugin`
- `just test -p codex-core-plugins`
- `just test -p codex-mcp-extension`
- `git diff --check`
## Why
Som tools, such as history and notes, must remain top-level when MCP
deferral is enabled while staying unavailable through code-mode `exec`.
## What changed
- Added `features.code_mode.direct_only_tool_namespaces`.
- Classified matching MCP tools as `DirectModelOnly`.
- Kept those tools top-level in `code_mode_only`.
- Excluded them from `tool_search` deferral and the nested `exec`
surface.
- Updated the generated config schema.
## Validation
- `code_mode_only_exposes_direct_model_only_mcp_namespaces`
- `load_config_resolves_code_mode_config`
Shijie RaoandGitHub
sayan-oaiandGitHub
xl-openaiandGitHub
pakrym-oaiandGitHub
rka-oaiandGitHub
Peter BakkumandGitHub
Alex DaleyandGitHub
Adam Perry @ OpenAIandGitHub
charlesgong-openaiandGitHub
AdrianandGitHub
canvrno-oaiandGitHub
Eric NingandGitHub
maja-openaiandGitHub
Matthew ZengandGitHub
TomandGitHub
Gabriel PealandGitHub
iceweasel-oaiandGitHub
Eric TrautandGitHub
jifandGitHub
Won ParkandGitHub