OwnMem CLI
Git-native project memory for AI agents: Markdown inside the repo, local recall with no model call, and gated delivery.
- 426 ★ on the repository
- Last commit
- Apache-2.0
Extended access
Why this level
- `init` adds hooks that run before agent edits.
- Memory writes files under `.ownmem/` inside the repository.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add cli/ownmemDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- `init` adds hooks that run before agent edits.
- Memory writes files under `.ownmem/` inside the repository.
This is third-party code. Review the repository files before installing.
What it does
OwnMem keeps project knowledge in `.ownmem/` as plain Markdown files. They go through review and rollback with the code, and every agent in the repository reads them: Claude Code, Codex, Cursor, Gemini CLI and Grok CLI. `ownmem recall` runs locally with no model call and no network. The answer arrives in one of three forms: a quoted memory, up to three pointers, or an abstention that names the gate that refused. The agent drafts a new memory, and a person commits it.
Who it is for. Teams where several agents work in one repository and repeat the same mistakes after losing context.
Good fit when
- Staging deploys keep failing for one reason and an agent guesses again each time.
- Project knowledge should be reviewed like code rather than kept in a vendor account.
- Claude Code and Codex both work in the repository and need shared memory.
- Before a change you want to ask whether the same failure was seen before.
Not a fit when
- You only need working rules, which `CLAUDE.md` or `AGENTS.md` already cover.
- Memory should fill automatically with no human review.
- The project cannot use Node.js 20.6 or newer.
Example request
Before adding workers, check through OwnMem whether the staging deploy timeout has been seen before.Limitations
Node.js 20.6 or newer is required. Commands install inside the project: `npm install --save-dev ownmem`, then `npx ownmem init --hook --hosts claude,codex` adds hooks before agent edits. Codex hooks need three one-time trust steps that `init` prints. For Cursor and Gemini CLI the MCP server is connected by hand. When upgrading from 0.6.0 run `npx ownmem init --update` first, because old hooks call subcommands that no longer exist. Memory does not grow without a human commit, so knowledge builds up slower than with auto-writes.
How to disable. Remove the hooks added by `init`, drop `ownmem` from `devDependencies` and delete `.ownmem/` if the memory is no longer needed.
FAQ
Where is memory stored?
In `.ownmem/` inside the repository as plain Markdown. It travels with clones, reviews and rollbacks.
Does recall call a model?
No. Recall by default uses no model and no network, and the same query gives the same ranking.
How do I create a new memory?
Run `npx ownmem new <name>`: it scaffolds one memory that already passes every gate. Then commit it like ordinary code.
What about hosts without hooks?
Run `npx ownmem mcp`: the server gives the agent two tools, `recall` and `read`. Neither can change memory.
Where do I see usage stats?
`npx ownmem report --since 7d` prints a summary and `npx ownmem dashboard --open` opens the local console.
Discussion
No questions or reviews yet.
Related
CLI toolsLocal memory for AI agents in plain Markdown files: it keeps only what you approve and serves it through MCP
Long-term memory for AI agents: Markdown files as the source of truth, a SQLite index as a cache, shared by Claude Code, Codex and others
Shared local memory for AI agents: Claude Code, Codex, Cursor and others find the agreements of past sessions in it and record new ones
Local memory for agents with a review queue: SQLite, an audit log and a token ledger, one vault for Claude Code, Codex and Cursor
Long-term agent memory in local Markdown files with a link graph, semantic search and MCP access
ai-memory retains session knowledge in a Markdown wiki, exposes MCP retrieval and passes context between agents and machines.
Comparison with similar tools
Cards that share the main search query, and how they differ.
| Tool | Official | Repository stars | Main difference |
|---|---|---|---|
| OwnMem (this card) | no | 426 | |
| Episodic Memory | no | 499 | searches past conversations by meaning; OwnMem keeps reviewable Markdown memories in the repository |
| claude-supermemory | yes | 2,803 | Supermemory long-term memory across sessions; OwnMem lives in Git with no external service |
README in short
Memory has three tiers: routing, area indexes and topics. Before answering, relevance, validity, task fit and action risk are checked.