deepagents
https://github.com/langchain-ai/deepagents
Python
Deep Agents is an agent harness built on langchain and langgraph. Deep Agents are equipped with a planning tool, a filesystem backend, and the ability to spawn subagents - making them well-equipped to handle complex agentic tasks.
Triage Issues!
When you volunteer to triage issues, you'll receive an email each day with a link to an open issue that needs help in this project. You'll also receive instructions on how to triage issues.
Triage Docs!
Receive a documented method or class from your favorite GitHub repos in your inbox every day. If you're really pro, receive undocumented methods or classes and supercharge your commit history.
Python not yet supported1 Subscribers
Add a CodeTriage badge to deepagents
Help out
- Issues
- Public API to replace/unregister a registered `HarnessProfile`
- LSP-backed post-edit diagnostics
- MCP tool-search / deferred-loading design
- `LocalShellBackend(virtual_mode=True)` silently nests host paths returned by `execute` when reused by file tools
- LLM Safety Classifier in `HumanInTheLoopMiddleware` for Auto Tool Execution
- Add `SANDBOX_TYPE=docker` for local containerized development/testing
- `SkillsMiddleware` never re-scans skill sources on a checkpointed thread: added skills stay invisible, deleted skills stay advertised
- cron ticker dies on store errors and stop re-raises the task failure
- `ThreadWorker.close()` and `ThreadWorker.run_sync()` block the asyncio event loop
- `FilesystemMiddleware` has no `permissions` parameter, so the documented `tools=` override silently drops `create_deep_agent(permissions=...)`
- Docs
- Python not yet supported