Project-M open-source local shared memory MCP server
Understand this story
Developer adityadebnath released the open-source project Project-M, a local-first MCP server for sharing project state across Claude Desktop, OpenAI Codex, and Google Antigravity, with the goal of avoiding context loss when switching tools. According to the author, Project-M runs entirely locally, uses SQLite WAL mode, incurs no API costs, and requires no paid embedding sync or third-party vector database; the project is open-sourced under the MIT license, and users can use it without cloning the repository.
AI overview based on source reports · Updated 19 hr ago
Coverage timeline
Follow the coverage to explore different perspectives.
- Hacker News AIShow HN: Project-M – Stop re-explaining project context between AI coding tools
Project-M is a local-first Model Context Protocol server that harvests session turns from Claude Desktop, OpenAI Codex, and Google Antigravity into a SQLite WAL database and compiles a deterministic Tier 1 briefing card of under 450 tokens in under 2 milliseconds. It maintains an immutable ADR ledger, a cross-agent task baton graph, and lossless raw transcripts, with SQLite FTS5 BM25 search and eight FastMCP tools exposed over stdio. The project is MIT-licensed and installable via uvx or a one-click setup command that injects the server into each client's config.
Activity for this story
- Current comparable activity
- 3
- Peak comparable activity
- 10Nov 18 23:00
- Change in the past 24 hours
- -49%
Trends compare the same fully observed participants, which may cover fewer participants than current activity. Move the pointer or tap the chart to inspect hourly activity; use the left and right arrow keys with a keyboard.