Ship a GitHub PR From a Slack Message with Managed Deep Agents

Summary

This walkthrough introduces Patch, an agent built using LangChain's Managed Deep Agents (MDAs). Patch automates the process of converting natural language feature requests from a Slack message into a fully drafted GitHub Pull Request (PR), complete with descriptions and code changes. The agent successfully demonstrated implementing features (e.g., adding a share button, changing the background color) for a Tetris side project, proving that complex integrations—including Slack communication, GitHub interaction, and code sandboxing—can be achieved with minimal code.

Technical details

• Agent Definition (agent.py) 151s

  The agent is defined in `agent.py`, specifying its name (Patch) and the underlying model (Claude Sonnet 5).

• Agent Instructions (instructions.md) 172s

  The `instructions.md` file guides the agent, pointing it to the target GitHub repository and instructing it to solve bugs and implement features, while also detailing the required behavior and procedure.

• GitHub Integration

  GitHub access is managed using the GitHub MCP (Managed Connector Platform) and requires an access token, which is stored as a secret environment variable (e.g., `Patch GitHub`).

• Sandbox Implementation

  The agent's ability to write and test code is enabled by importing and invoking the `define_sandbox` function, providing a secure environment for code execution.

Mentioned resources

• Managed Deep Agents docs (Documentation)
• Managed Deep Agents: the fastest way to ship a production deep agent (Blog Post)
• Connect a Managed Deep Agent to Slack (Documentation)
• LangSmith (Platform)

Channel & topics

• LangChain
• Agentic Workflow
• LLM Orchestration
• GitHub API Integration
• Slack API
• Code Sandboxing
• LangChain
• Managed Deep Agents docs
• Managed Deep Agents: the fastest way to ship a production deep agent
• Connect a Managed Deep Agent to Slack
• LangSmith