potpie: what it is, what problem it solves & why it's gaining traction

What it solves

Potpie addresses the lack of project-specific context for AI agents during software development. It prevents agents from guessing or lacking knowledge about a codebase's structure, historical decisions, and team workflows, enabling them to answer questions, plan changes, and write code more accurately.

How it works

Potpie creates a "living context graph" by indexing multiple sources of engineering data, including code, source history, team knowledge, and software development lifecycle (SDLC) data. It provides a CLI and a web UI for managing this graph, and integrates directly into coding harnesses (like Cursor or Claude Code) by installing specific instructions and skills that allow agents to pull relevant context automatically.

Who it’s for

Developers and engineering teams who use AI agents to assist with coding, debugging, and project planning and want those agents to have a deep, integrated understanding of their specific codebase and organizational knowledge.

Highlights

• Multi-source indexing: Integrates with GitHub, Linear, Jira, and Confluence to capture code, issues, and documentation.
• Agent-ready skills: Provides specific guidance and skills for popular AI coding tools like Claude Code, OpenAI Codex, Cursor, and OpenCode.
• Context resolution: Includes a resolve command to pull the exact context an agent needs before starting a specific task.
• Graph visualization: Includes a web UI to explore the project's context graph visually.