PAPI

What is PAPI?

PAPI is a Model Context Protocol server that gives AI assistants structured project memory—plans, builds, reviews, decisions—that persists across sessions, tools, and teammates. It is designed for developers and teams who want to keep AI coding tools aware of the project’s current cycle, status, and next steps.

How to use PAPI?

Connect PAPI to your AI tool as an MCP server. The fastest path is Claude Code: run claude mcp add --transport http papi https://mcp.getpapi.ai/mcp. A browser tab opens for sign‑in, then you can run the setup tool to scaffold the project and orient to discover the current cycle. For other tools (Cursor, VS Code, Windsurf, Codex), see docs/install.md.

Key features of PAPI

• Plan, build, review, and release cycles that persist across sessions.
• Strategy reviews every few cycles to check direction.
• Dashboard at getpapi.ai showing cycles, board, and decisions.
• Memory that compounds—assistant starts every session knowing the project state.
• Methodology embedded in the product: plan, build, review, release loop.

Use cases of PAPI

• Give an AI assistant persistent project context so it never starts from zero.
• Track what was built, what surprised you, and what was discovered in each cycle.
• Close the loop between cycles so every plan feeds the next.
• Maintain a shared project memory across multiple AI tools and teammates.

FAQ from PAPI

What does PAPI do that a simple prompt or notes file can’t?

PAPI provides structured, persistent project memory—plans, builds, reviews, and decisions—that your AI assistant reads and writes automatically. It keeps the project state across sessions, tools, and teammates, so you don’t have to re‑explain everything.

What are the runtime or dependency requirements?

PAPI is an MCP server. The supported local runtime is the @papi-ai/server package on npm. The PAPI engine itself is closed source; this repository is documentation only.

Where is my project data stored?

PAPI uses a cloud dashboard at getpapi.ai to show cycles, board, and decisions. The underlying engine is closed source, so data likely lives on PAPI’s infrastructure. For local runs, the @papi-ai/server package may handle data locally—consult the documentation.

What transport protocols and authentication does PAPI support?

The quick start uses HTTP transport (--transport http). Authentication is handled via a browser sign‑in flow. No other transports or auth methods are mentioned in the README.

Is PAPI open source?

No. The PAPI engine is closed source. The repository contains only documentation, and the @papi-ai/server npm package is the supported local runtime.