Set up Memmo

npm install -g @memmo/cli
memmo login

login opens your browser to authorize the machine and stores a token in ~/.memmo/credentials.json (--token for CI). On first sign-in you name your project and the app walks you through connecting a repo — three short steps.

From the repo root:

memmo init

It registers the repo (from its git remote) in a project, then writes:

.mcp.json                 # the Memmo MCP server (runs via npx — nothing to install)
CLAUDE.md                 # a short section telling the agent how to use memory
.claude/settings.json     # hooks: SessionStart · UserPromptSubmit · PreToolUse · SessionEnd

Commit these three. Teammates who clone the repo only need step 1. Then say yes to the scan: Claude Code reads the codebase and proposes starter memories to your inbox (a couple of minutes, runs on your Claude subscription).

Every Claude Code session in that repo now:

• starts with a briefing — rules, architecture, key commands, known risks;
• recalls what matters for the task when you type a prompt (semantic search);
• is warned before editing risky files (memories with file globs);
• proposes new memories when it ends — including corrections to outdated ones.

The agent can also call the tools directly:

get_repo_context()          → the briefing
search_memory(query)        → approved memories, ranked by meaning
get_relevant_warnings(files)→ risks for the files about to change
propose_memories([...])     → record something worth keeping (goes to the inbox)

The app has four screens. Inbox is where proposals wait for a human — approve, edit, reject, or set confidence thresholds to auto-triage. Memory is everything approved, searchable by meaning, with a “New memory” button for the things you already know.

Any MCP client can read the same memory over HTTP. Generate a connector command under Connect in the app, or configure it by hand:

claude mcp add --transport http memmo https://api.memmo.dev/mcp \
  --header "Authorization: Bearer <token>"

The agent passes its repo (git remote or owner/repo) per call and Memmo routes it to the right project. Hooks (automatic briefing and capture) are Claude Code-only; other agents get the tools.

A project is the set of repos that share one memory — one monorepo, or many services. Each memory applies to every repo in the project or to specific ones. An organization holds projects, members and billing; belonging to it is enough to use its projects.