Add comprehensive documentation for the SSE stream termination fix:
- Problem analysis and root cause
- Step-by-step solution with code examples
- Security considerations (__debug__ vulnerability)
- Code simplification recommendations
- Prevention strategies and best practices
- Testing and monitoring guidelines
Location: docs/solutions/runtime-errors/sse-mcp-tool-error-handling.md
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Remove __debug__ conditional that exposed tracebacks in production
- Simplify error response structure to {"error": str(e)}
- Remove redundant exception handling in init_agent()
- Remove unnecessary traceback logging
- Reduce code from ~30 lines to ~10 lines
Fixes security vulnerability where __debug__ is always True
unless Python runs with -O flag, causing full tracebacks
to be sent to clients in production.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Create new skill_manager.py with list and upload endpoints
- Add GET /api/v1/skill/list to retrieve official and user skills
- Add POST /api/v1/skill/upload for skill file upload
- Parse SKILL.md frontmatter to extract name and description
- Move skill upload endpoint from files.py to skill_manager.py
- Add SKILLS_DIR configuration to settings.py
- Register skill_manager router in fastapi_app.py
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Enhance exception handling in agent_task() to capture and send structured error messages via SSE stream
- Add [DONE] marker to outer exception handler to ensure proper stream termination
- Improve MCP tool loading error handling in init_agent() to prevent cascading failures
- Add detailed error logging with traceback for debugging
Fixes RemoteProtocolError that occurred when MCP tool calls failed,
which previously caused incomplete chunked read errors and connection drops.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Refactor _extract_skills_to_robot to accept bot_id instead of robot_dir
- Add multi-directory skill search with priority order
- Switch from zip extraction to direct directory copying
- Add rag-retrieve skill directory
Add new POST endpoint /api/v1/skills/upload for uploading skill zip files.
The endpoint:
- Accepts zip files with bot_id parameter
- Validates file format (must be .zip)
- Saves zip to projects/uploads/{bot_id}/skill_zip/
- Automatically extracts to projects/uploads/{bot_id}/skills/{skill_name}/
- Returns success response with file and extract paths
This enables programmatic skill deployment for specific bots.
- Change workspace_root from ~/.deepagents/{bot_id} to projects/robot/{bot_id}
- Refactor CustomSkillsMiddleware to support user and project skills separately
- Update skill path formatting for better visibility in system prompts
- Remove unused symlink_utils module and related setup code
- Clean up imports in fastapi_app.py and utils/__init__.py
Remove the CRITICAL section about directory access restrictions.
Path handling guidance is still maintained.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Override AgentMemoryMiddleware to simplify agent directory display.
The CustomAgentMemoryMiddleware shows "." instead of full path.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add CustomSkillsMiddleware to show ./skills instead of full paths
- Enable virtual_mode for FilesystemBackend in local mode
- Display "." instead of ~/.deepagents/{bot_id} in system prompt
- Remove backend_host and masterkey from robot project config
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add create_custom_cli_agent function to support custom workspace_root
- Set shell workspace to ~/.deepagents/{bot_id} for deep_agent type
- Pass system_prompt to create_custom_cli_agent for proper context
- Fix duplicate <env> tag in system_prompt_deep_agent.md
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Set enable_shell=False to disable shell command execution capabilities
in the deep agent for security reasons.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
The create_folder endpoint was incorrectly defined to accept query
parameters instead of JSON request body, causing 400 errors when
called from the frontend.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
When skills parameter is empty or None and robot_type is
catalog_agent or deep_agent, automatically load rag-retrieve skill.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add skills parameter to ChatRequest for skill file processing
- Extract and unzip skill files to robot project skills directory
- Add robot_config.json with bot_id and environment variables
- Update symlink setup to skip if ~/.deepagents already exists
- Enhance system prompt with directory access restrictions
- Refactor _get_robot_dir to handle symlink paths correctly
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
