Enhance MCP development features and introduce skills management

- Added configuration options for requiring human approval before applying LLM-generated MCP patches.
- Updated Docker setup to include skills directory.
- Integrated skills management into the backend, allowing for procedural guides and skill matching.
- Refactored database initialization to apply Alembic migrations.
- Enhanced task approval process to handle MCP patch applications with optional approval.
- Introduced new schemas for skills and updated existing APIs to support skills functionality.

This commit lays the groundwork for improved agent capabilities and better management of MCP development processes.

backend/app/api/tasks.py

@@ -351,7 +351,29 @@ async def decide_approval(
     approval.decided_by_id = user.id
     approval.decided_at = datetime.now(timezone.utc)
 
-    if body.approve and approval.tool_name == "mcp_git_push" and approval.tool_args:
+    if body.approve and approval.tool_name == "mcp_apply_patch" and approval.tool_args:
+        from app.mcp_manager.commands import enqueue_mcp_command
+
+        args = approval.tool_args
+        result = await enqueue_mcp_command(
+            "apply_patch",
+            server=args.get("server"),
+            timeout_seconds=180,
+            task_id=task_id,
+            tool_args=args,
+        )
+        if not result.get("ok"):
+            raise HTTPException(
+                status_code=502,
+                detail=result.get("error", "MCP patch apply failed"),
+            )
+        await publish_event(
+            task_id,
+            "mcp_dev",
+            result.get("summary") or f"MCP patch applied to {args.get('server')}",
+            {"kind": "mcp_apply_patch", **result},
+        )
+    elif body.approve and approval.tool_name == "mcp_git_push" and approval.tool_args:
         from app.services import mcp_workspace
 
         args = approval.tool_args