Python: Fix server_tool_use input_json_delta handling and improve Anthropic samples (#5050)

* Fix server_tool_use input_json_delta handling and improve Anthropic samples

- Fix: Skip input_json_delta for server_tool_use content blocks in AnthropicClient streaming. Server-managed tools (e.g., skills with code interpreter) were producing Content.from_function_call(name='') entries that caused Anthropic API 400 errors on subsequent turns.

- Samples: Add dotenv loading and environment variable documentation to Anthropic Claude samples (MCP, permissions, session, shell, tools, URL, skills).

* Add regression test for server_tool_use + input_json_delta skip behavior

Agent-Logs-Url: https://github.com/microsoft/agent-framework/sessions/7c68dcb2-b577-4e36-b423-664b8fe3ac1d

Co-authored-by: chetantoshniwal <255221507+chetantoshniwal@users.noreply.github.com>

---------

Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
Co-authored-by: chetantoshniwal <255221507+chetantoshniwal@users.noreply.github.com>

python/samples/02-agents/providers/anthropic/anthropic_claude_with_mcp.py

@@ -12,6 +12,9 @@ Supported MCP server types:
 - "http": Remote HTTP server
 - "sse": Remote SSE (Server-Sent Events) server
 
+Environment variables:
+- ANTHROPIC_API_KEY: Your Anthropic API key
+
 SECURITY NOTE: MCP servers can expose powerful capabilities. Only configure
 servers you trust. Use permission handlers to control what actions are allowed.
 """

python/samples/02-agents/providers/anthropic/anthropic_claude_with_multiple_permissions.py

@@ -15,6 +15,9 @@ Available built-in tools:
 - "Glob": Search for files by pattern
 - "Grep": Search file contents
 
+Environment variables:
+- ANTHROPIC_API_KEY: Your Anthropic API key
+
 SECURITY NOTE: Only enable permissions that are necessary for your use case.
 More permissions mean more potential for unintended actions.
 """
@@ -24,6 +27,10 @@ from typing import Any
 
 from agent_framework.anthropic import ClaudeAgent
 from claude_agent_sdk import PermissionResultAllow, PermissionResultDeny
+from dotenv import load_dotenv
+
+# Load environment variables from .env file
+load_dotenv()
 
 
 async def prompt_permission(

python/samples/02-agents/providers/anthropic/anthropic_claude_with_session.py

@@ -6,6 +6,9 @@ Claude Agent with Session Management
 This sample demonstrates session management with ClaudeAgent, showing
 persistent conversation capabilities. Sessions are automatically persisted
 by the Claude Code CLI.
+
+Environment variables:
+- ANTHROPIC_API_KEY: Your Anthropic API key
 """
 
 import asyncio
@@ -14,8 +17,12 @@ from typing import Annotated
 
 from agent_framework import tool
 from agent_framework.anthropic import ClaudeAgent
+from dotenv import load_dotenv
 from pydantic import Field
 
+# Load environment variables from .env file
+load_dotenv()
+
 
 @tool
 def get_weather(

python/samples/02-agents/providers/anthropic/anthropic_claude_with_shell.py

@@ -7,6 +7,9 @@ This sample demonstrates how to enable shell command execution with ClaudeAgent.
 By providing a permission handler via `can_use_tool`, the agent can execute
 shell commands to perform tasks like listing files, running scripts, or executing system commands.
 
+Environment variables:
+- ANTHROPIC_API_KEY: Your Anthropic API key
+
 SECURITY NOTE: Only enable shell permissions when you trust the agent's actions.
 Shell commands have full access to your system within the permissions of the running process.
 """
@@ -16,6 +19,10 @@ from typing import Any
 
 from agent_framework.anthropic import ClaudeAgent
 from claude_agent_sdk import PermissionResultAllow, PermissionResultDeny
+from dotenv import load_dotenv
+
+# Load environment variables from .env file
+load_dotenv()
 
 
 async def prompt_permission(

python/samples/02-agents/providers/anthropic/anthropic_claude_with_tools.py

@@ -13,11 +13,18 @@ Available built-in tools:
 - "Edit": Edit existing files
 - "Glob": Search for files by pattern
 - "Grep": Search file contents
+
+Environment variables:
+- ANTHROPIC_API_KEY: Your Anthropic API key
 """
 
 import asyncio
 
 from agent_framework.anthropic import ClaudeAgent
+from dotenv import load_dotenv
+
+# Load environment variables from .env file
+load_dotenv()
 
 
 async def main() -> None:

python/samples/02-agents/providers/anthropic/anthropic_claude_with_url.py

@@ -10,6 +10,9 @@ Available web tools:
 - "WebFetch": Fetch content from URLs
 - "WebSearch": Search the web
 
+Environment variables:
+- ANTHROPIC_API_KEY: Your Anthropic API key
+
 SECURITY NOTE: Only enable URL permissions when you trust the agent's actions.
 URL fetching allows the agent to access any URL accessible from your network.
 """
@@ -17,6 +20,10 @@ URL fetching allows the agent to access any URL accessible from your network.
 import asyncio
 
 from agent_framework.anthropic import ClaudeAgent
+from dotenv import load_dotenv
+
+# Load environment variables from .env file
+load_dotenv()
 
 
 async def main() -> None:

python/samples/02-agents/providers/anthropic/anthropic_skills.py

@@ -21,6 +21,10 @@ This sample demonstrates using Anthropic with:
     You can also set additonal_chat_options with "additional_beta_flags" per request.
 - Creating an agent with the Code Interpreter tool and a Skill.
 - Catching and downloading generated files from the agent.
+
+Environment variables:
+- ANTHROPIC_API_KEY: Your Anthropic API key
+- ANTHROPIC_CHAT_MODEL_ID: The Anthropic model to use, such as "claude-sonnet-4-5-20250929"
 """