* Implement MCP background tasks (SEP-1686) using Docket
Adds support for background task execution via the MCP task protocol,
powered by Docket for task queue management.
- Tools, resources, and prompts can be marked with `task=True` to run async
- Progress dependency for tracking task progress
- CurrentDocket and CurrentWorker dependencies for advanced use cases
- Client API with `.call_tool(..., task=True)` returns task handles
- Task status notifications via subscriptions
- CLI worker command for distributed task processing
Configuration via environment:
- FASTMCP_ENABLE_DOCKET=true
- FASTMCP_ENABLE_TASKS=true
- FASTMCP_DOCKET_URL=redis://... (or memory:// for single-process)
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Fix tasks example import (TaskStatusResponse → GetTaskResult)
The example was using a non-existent TaskStatusResponse type.
Updated to use mcp.types.GetTaskResult which is what the
on_status_change callback actually receives.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Fix env var name in Docket error messages
The error messages referenced FASTMCP_EXPERIMENTAL_ENABLE_DOCKET but the
actual setting is FASTMCP_ENABLE_DOCKET.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Remove deprecated code re-added from pre-#2329 branch
- Remove ExtendedEnvSettingsSource (FASTMCP_SERVER_ prefix support)
- Remove dependencies parameter from FastMCP.__init__
* Replace fakeredis git pin with PyPI release
* Remove redundant fakeredis dev dep (pulled via pydocket)
---------
Co-authored-by: Claude <noreply@anthropic.com>
Co-authored-by: Jeremiah Lowin <153965+jlowin@users.noreply.github.com>
* feat: Add optional meta parameter to Client tool call methods
* fix: Add support for mcp<1.19
* chore: cleaner solution
* Refactor call_tool to directly accept meta parameter and add tests for meta functionality
---------
Co-authored-by: Jeremiah Lowin <153965+jlowin@users.noreply.github.com>
* Add manual initialization control to Client
- Add auto_initialize parameter (default True) to control automatic initialization
- Make initialize() method public with idempotent caching
- Add comprehensive test suite for initialization behavior
* Document client initialization control and server instructions
- Expand documentation to cover auto_initialize parameter
- Show manual initialization for advanced use cases
- Document accessing server instructions via initialize_result
* Update client.mdx
* Refactor Client session state with ClientSessionState dataclass
Fixes#1068 by introducing ClientSessionState to encapsulate session management
attributes, simplifying Client.new() and preventing concurrent proxy client
context mixing through client factory pattern.
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Update proxy documentation for new client factory pattern
Documents the new session management behavior, client_factory parameter,
and concurrent operation safety introduced in v2.10.3 to fix#1068.
* Remove deprecation of client parameter and update documentation
- Undeprecated FastMCPProxy client parameter
- Made client_factory the advanced option for custom control
- Added detailed explanation of how client factories work internally
- Removed outdated note about proxy feature limitations
- Removed fabricated Advanced Usage section
* Re-do proxy documentation
* Fix deprecated client parameter to provide session isolation
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
---------
Co-authored-by: Claude <noreply@anthropic.com>
Single-element lists like [1] were being incorrectly unwrapped to "1"
in unstructured content while multi-element lists remained as lists.
This created inconsistent behavior where the structure was lost for
single items.
This fix ensures lists always preserve their structure in unstructured
content regardless of length, making behavior consistent and predictable.
Also removes pretty-printing from JSON serialization for more compact
output across tools, prompts, and resources.
Fixes#1064🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Add pydantic_core.to_json() serialization for non-string prompt arguments
- Update type annotations to accept dict[str, Any] instead of dict[str, str]
- Add focused tests covering specific scenarios:
* Client always serializes non-string args regardless of server types
* Integration with server-side type conversion
* Client serialization error with specific PydanticSerializationError
* Server deserialization error with specific McpError match
This ensures MCP protocol compliance while maintaining developer experience
with typed arguments.
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
