potpie
Context Graph for AI Native SDLC
Potpie is an agent skill that transforms codebases and software development workflows into a queryable context graph for AI coding assistants. It indexes code structure, commit history, pull requests, issues, team decisions, and engineering documentation from sources like GitHub, Linear, Jira, and Confluence. AI agents working in harnesses such as Claude Code, Cursor, OpenAI Codex, or OpenCode can use Potpie's CLI to resolve task-relevant context, search for specific patterns or decisions, and record new project knowledge. The skill helps agents answer questions, plan changes, and write code with accurate, repository-specific understanding.
Key Features
Use Cases
- 01Help AI coding agents understand repository architecture and conventions before generating code
- 02Enable agents to search for authentication flows, design decisions, or coding patterns in project history
- 03Let agents discover related pull requests, issues, and reviews when debugging or planning features
- 04Provide agents with team knowledge from Confluence runbooks and Linear project documentation
- 05Record agent decisions or discovered patterns back into the context graph for future reference
- 06Verify that agents have up-to-date context for multi-repository or polyglot projects
Related Skills
View moresuperpowers
An agentic skills framework & software development methodology that works.
skills
Skills for Real Engineers. Straight from my .agents directory.
skills
Public repository for Agent Skills
ponytail
Makes your AI agent think like the laziest senior dev in the room. The best code is the code you never wrote.
potpie — FAQ
What is Potpie?+
Potpie is an agent skill that indexes your codebase, development history, issues, pull requests, and team documentation into a context graph. AI coding agents query this graph via CLI commands to get project-specific context for tasks like code generation, debugging, and planning.
How do I install Potpie?+
Install the Potpie CLI using 'uv tool install potpie' or 'pip install --user potpie', then run 'potpie setup' to configure storage, the daemon, and agent skill integration for your chosen coding harness. The setup wizard will guide you through selecting integrations and the AI client you want to use.
Which AI coding clients work with Potpie?+
Potpie installs skills for Claude Code, OpenAI Codex, Cursor, and OpenCode. The setup wizard lets you choose your harness, and Potpie will configure the necessary instructions and CLI hooks for that environment.
Do I need API keys or authentication for Potpie?+
You need to authenticate integrations like GitHub, Linear, Jira, or Confluence using commands like 'potpie github login' so Potpie can index their data. The CLI includes 'potpie auth status' to verify your credentials and check integration health.
Is Potpie free to use?+
Potpie is open source under the Apache 2.0 license. The CLI and local context graph are free; the README mentions optional account-backed features via 'potpie login' but does not specify pricing.
Do I need to manually ingest my repository?+
No. After you register a repository with 'potpie source add repo .', the daemon and agent can ingest or update context automatically when tasks require it. You can monitor readiness with 'potpie status'.
How do I install potpie?+
Open the source repository on GitHub and follow its README. potpie is a skill — MCP Agents Market links you directly to the official repo.
Is potpie free?+
potpie is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.