Repolode
Submit
potpie-ai

potpie-ai/potpie

A CLI-driven context graph that indexes your codebase, issues, and team knowledge so AI coding agents work with project-specific context.

View on GitHubWebsite
Stars
5,748
Forks
686
Open issues
113
Last commit
6 hours ago

What is it?

Potpie builds a living context graph from a codebase and its software development lifecycle, covering code structure, decisions, source history, team knowledge, and engineering workflows. Coding agents can query this graph to answer questions, plan changes, debug failures, and write code that fits the project.

The project is CLI-first and designed for both humans and agents. A setup wizard provisions local config, storage, a daemon, a default workspace (called a pot), and agent skills. Commands such as resolve and search pull relevant context before a task, while record stores durable project learnings. A local web UI provides a graph explorer.

Integrations currently include GitHub, Linear, Jira, and Confluence. Supported coding harnesses are Claude Code, OpenAI Codex, Cursor, and OpenCode. The project is distributed on PyPI and licensed under Apache 2.0.

Key features

  • 01Indexes code structure, pull requests, issues, and source history into one context graph.
  • 02Provides resolve and search commands to fetch task-relevant context for agents.
  • 03Lets agents record durable project decisions and learnings with the record command.
  • 04Connects to GitHub, Linear, Jira, and Confluence sources.
  • 05Installs agent skills for Claude Code, Codex, Cursor, and OpenCode.

Use cases

  • Giving a coding agent project background before it plans or edits code in an unfamiliar repository.
  • Looking up where an authentication flow, bug, or convention lives across a codebase.
  • Capturing architectural decisions so future agent sessions can reference them.
  • Debugging failures by linking code history with related issues and documents.

Quick start

  1. 1

    Install the Potpie CLI from PyPI with uv

    uv tool install potpie
  2. 2

    Run the Potpie setup wizard

    potpie setup
  3. 3

    Open your configured coding harness and ask it to use Potpie for the repo

Things to know

  • Requires signing in to Potpie for account-backed and managed features.
  • Integrations such as GitHub, Linear, Jira and Confluence need separate credentials to connect.

Topics

#ai-agents#knowledge-graph#context-engineering#developer-tools#code-search#rag#cli#agents#ai-agents-framework#artificial-intelligence

FAQ

What is potpie?
A CLI-driven context graph that indexes your codebase, issues, and team knowledge so AI coding agents work with project-specific context. Potpie builds a living context graph from a codebase and its software development lifecycle, covering code structure, decisions, source history, team knowledge, and engineering workflows. Coding agents can query this graph to answer questions, plan changes, debug failures, and write code that fits the project.
Is potpie free to use? What license does it have?
potpie is open source under the Apache-2.0 license. Check the license terms before using it in commercial products.
Is potpie still maintained?
Actively maintained. The last commit to potpie was on Oct 22, 2026.
What can I use potpie for?
Giving a coding agent project background before it plans or edits code in an unfamiliar repository. Looking up where an authentication flow, bug, or convention lives across a codebase. Capturing architectural decisions so future agent sessions can reference them. Debugging failures by linking code history with related issues and documents.
How do I install potpie?
Install the Potpie CLI from PyPI with uv: uv tool install potpie Run the Potpie setup wizard: potpie setup Open your configured coding harness and ask it to use Potpie for the repo
Where can I get potpie?
The source code and installation instructions are on GitHub: https://github.com/potpie-ai/potpie

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects