Aari 47b258ebd7
feat(mcp): add ordinary durable task driver (#4690)
* feat(mcp): add durable task runtime foundation

* fix(chart): sync embedded config version

* fix(mcp): isolate task polls during shutdown

* feat(mcp): track consecutive poll errors on mcp_tasks

poll_attempt_count grows on every claim (successful polls included), so it
cannot drive a failure backoff without misjudging normal long tasks. Add
consecutive_poll_error_count: incremented when a claim is released after a
poll error, reset to zero by any applied snapshot. The backoff/terminal
policy that consumes it lands with the first concrete driver.

* fix(mcp): harden durable task lifecycle

* feat(mcp): add ordinary durable task driver

* test(mcp): address durable task review feedback

* fix(mcp): preserve submit tool descriptions

* fix(mcp): bound remote task calls

* fix(mcp): bound persisted task payloads

* fix(mcp): preserve task tool error details

* fix(mcp): enforce durable task boundaries

* test(mcp): cover task config snapshot lifecycle

---------

Co-authored-by: Willem Jiang <willem.jiang@gmail.com>
2026-08-15 14:26:38 +08:00

96 lines
3.1 KiB
Python

"""Thread-scoped read API for durable MCP background tasks."""
from __future__ import annotations
from typing import Any
from fastapi import APIRouter, HTTPException, Query, Request
from app.gateway.authz import require_permission
from app.gateway.deps import get_current_user, get_mcp_task_repo, get_mcp_task_service
from deerflow.utils.thread_id import ThreadId
router = APIRouter(prefix="/api/threads/{thread_id}/mcp-tasks", tags=["mcp-tasks"])
_MAX_PUBLIC_ERROR_CHARS = 500
def _short_error(value: Any) -> str | None:
if value is None:
return None
return str(value)[:_MAX_PUBLIC_ERROR_CHARS]
def _tracking_degraded(record: dict[str, Any], *, threshold: int) -> bool:
return int(record.get("consecutive_poll_error_count") or 0) >= threshold
def _list_item(record: dict[str, Any], *, threshold: int) -> dict[str, Any]:
return {
"task_id": record["id"],
"task_name": record["task_name"],
"status": record["status"],
"created_at": record["created_at"],
"updated_at": record["updated_at"],
"error": _short_error(record.get("error")),
"tracking_degraded": _tracking_degraded(record, threshold=threshold),
}
def _detail(record: dict[str, Any], *, threshold: int) -> dict[str, Any]:
return {
**_list_item(record, threshold=threshold),
"last_polled_at": record.get("last_polled_at"),
"last_poll_error": _short_error(record.get("last_poll_error")),
"result": record.get("result"),
"result_preview": record.get("result_preview"),
"result_truncated": bool(record.get("result_truncated")),
"result_artifact": record.get("result_artifact"),
"input_required": record.get("input_required"),
}
async def _current_user_id(request: Request) -> str:
user_id = await get_current_user(request)
if user_id is None:
raise HTTPException(status_code=401, detail="Authentication required")
return user_id
@router.get("")
@require_permission("threads", "read", owner_check=True)
async def list_mcp_tasks(
thread_id: ThreadId,
request: Request,
limit: int = Query(default=50, ge=1, le=100),
) -> list[dict[str, Any]]:
repository = get_mcp_task_repo(request)
service = get_mcp_task_service(request)
user_id = await _current_user_id(request)
records = await repository.list_by_thread(
thread_id,
user_id=user_id,
limit=limit,
)
threshold = service.tracking_degraded_after_errors
return [_list_item(record, threshold=threshold) for record in records]
@router.get("/{task_id}")
@require_permission("threads", "read", owner_check=True)
async def get_mcp_task(
thread_id: ThreadId,
task_id: str,
request: Request,
) -> dict[str, Any]:
repository = get_mcp_task_repo(request)
service = get_mcp_task_service(request)
user_id = await _current_user_id(request)
record = await repository.get(task_id, user_id=user_id)
if record is None or record["thread_id"] != thread_id:
raise HTTPException(status_code=404, detail="MCP task not found")
return _detail(
record,
threshold=service.tracking_degraded_after_errors,
)