# mex-memory/mex

Persistent project memory for AI coding agents. Structured scaffold + drift detection CLI.

Repository: https://github.com/mex-memory/mex
Canonical: https://ross.abutalabs.com/products/mex
Homepage: https://mexmemory.com
Language: TypeScript
License: MIT
License Family: permissive
Topics: claude-code, claude-code-skills, cli-tool, context-management, cursor, developer-tools, documentation, llm, codex, memory-management, typescript, mcp-server, code-graph, knowledge-graph, tree-sitter
Last push: 2026-08-26T19:33:56+00:00

## Health v2 (maintenance only)
Score: 81/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 98, longevity 11
- inputs: {"age_days": 165, "days_push": 7, "days_rel": 13, "gap_med": 15, "n_releases_24m": 10}
- flags: young
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1526, forks 107 (observed 2026-08-28T04:04:58.333476+00:00)

## What it is
mex is a TypeScript CLI that gives AI coding agents persistent, structured project memory: it maps the codebase into a deterministic code graph, turns agent learnings into Markdown wiki pages, and routes only task-relevant context into each session. It also offers an MCP server and drift detection that flags knowledge affected by code changes.

## Use cases
- give my AI coding agent persistent memory of the codebase
- stop agents re-scanning the whole repo every session
- detect when documentation drifts from the code
- share a common knowledge base across Claude Code, Cursor, and Codex
- build a code knowledge graph for my repository
- load only relevant architectural context per task
- set up an MCP server for project context

## When to choose
- you use AI coding agents (Claude Code, Cursor, Codex) on a large codebase and want structured, source-linked memory instead of a giant stale instruction file
- you want drift detection tying documentation to exact code symbols
- you want an MCP-compatible context server with deterministic graph retrieval

## When to avoid
- you need a fully managed team-hosted memory service - that's a separate early-access offering, not this open-source CLI
- your project is small enough that a single AGENTS.md file suffices
- you're not using AI coding agents at all

## Facets
- artifact type: cli-tool
- maturity: active
- function: developer-tools, documentation, search-engine, mcp, cli, parser
- domain: developer-tools, large-language-models, documentation, apis
- platform: cli, cross-platform
- tags: ai-coding-agents, agent-memory, context-management, code-graph, knowledge-graph, tree-sitter, claude-code, cursor, drift-detection, living-wiki, ai-agents, nodejs

## Member repositories
- mex-memory/mex (main) score 81

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:58.333476+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T04:31:37.254747+00:00, confidence not recorded.
  - readme: https://github.com/mex-memory/mex (fetched 2026-08-28T04:04:58.333476+00:00, sha 7526237ce96e)
  - homepage: https://mexmemory.com (fetched 2026-08-29T11:34:31.058532+00:00, sha d89f6caaadc7)
- Data as of 2026-08-30T08:39:29.467469+00:00.