potpie-ai/potpie
Context Graph for AI Native SDLC
What it solves
AI agents often lack the deep, project-specific context needed to perform complex software engineering tasks. Potpie solves this by indexing code, structure, decisions, source history, and team knowledge into a living context graph, allowing agents to answer questions, plan changes, and debug failures with high accuracy.
How it works
Potpie uses a CLI-first approach to index various data sources including local repositories, GitHub, Linear, Jira, and Confluence. It builds a context graph that can be explored via a web UI. The tool integrates with coding harnesses like Claude Code, Cursor, and OpenAI Codex, providing them with specialized skills and instructions to leverage the indexed project context.
Who it’s for
Software engineering teams and developers who want to empower their AI coding assistants with comprehensive, project-specific context and historical knowledge.
Highlights
- Indexes code, source history, and engineering workflows
- Integrates with GitHub, Linear, Jira, and Confluence
- Provides specialized skills for coding harnesses like Cursor and Claude Code
- Includes a web-based graph explorer for visualizing context