# Claude-Mem review

> Claude-Mem is a persistent memory plugin that captures coding agent sessions, compresses them with AI, and injects relevant context into future sessions.

Devtoolwatch Board, the week of Oct 25–31, 2026. Canonical page: https://index.devtoolwatch.com/tools/thedotmack-claude-mem/

## TL;DR

Claude-Mem is a memory system built for Claude Code that also works with Codex, OpenCode, OpenClaw and other agents. Claude-Mem captures tool usage, writes semantic summaries, and recalls them in later sessions. It is for developers who want agents to keep project knowledge between sessions, and it uses layered search to limit token use.

## Claude-Mem at a glance

- **Category:** [Coding agents](https://index.devtoolwatch.com/categories/coding-agents/)
- **Maintained by:** Alex Newman (individual, [@thedotmack](https://github.com/thedotmack) on GitHub)
- **License:** Apache 2.0
- **Language:** TypeScript, JavaScript, Python
- **Repository:** [github.com/thedotmack/claude-mem](https://github.com/thedotmack/claude-mem)
- **Website:** [claude-mem.ai](https://claude-mem.ai)
- **npm package:** [claude-mem](https://www.npmjs.com/package/claude-mem) 15.0.1
- **Latest release:** [v15.0.1](https://github.com/thedotmack/claude-mem/releases/tag/v15.0.1) on GitHub, Oct 20, 2026
- **First published:** Aug 31, 2025

## What Claude-Mem does

- Claude-Mem captures tool usage observations during agent sessions and generates semantic summaries for future sessions.
- Claude-Mem injects relevant context from previous sessions into new sessions automatically.
- Claude-Mem provides MCP search tools: search, timeline, and get_observations, used in a layered workflow that fetches full details only for selected IDs.
- Claude-Mem uses Smart search by default, where the agent reasons about relevance between steps. A faster dumb_search does deterministic title matching.
- Claude-Mem offers a local My Work view of requests given to agents, with cards, list and per-agent views.
- Claude-Mem lets you exclude sensitive content from storage using privacy tags.
- Claude-Mem supports an optional TeamBrain shared workspace where each person chooses what to share.

### How it works

Claude-Mem registers lifecycle hooks such as SessionStart, PostToolUse and SessionEnd, each running a claude-mem hook command. A local worker service managed by Bun provides an HTTP API, web viewer and search endpoints. Data is stored in SQLite with FTS5, and Chroma adds hybrid semantic and keyword search. Settings live in ~/.claude-mem/settings.json.

## How to install Claude-Mem

**With npx**

```
npx claude-mem install
```

**For OpenCode**

```
npx claude-mem install --ide opencode
```

**On an OpenClaw gateway**

```
curl -fsSL https://install.cmem.ai/openclaw.sh | bash
```

### Requirements

- Node.js 20.12.0 or higher (README states 20.0.0 or higher).
- Bun 1.1.31 or higher, auto-installed if missing.
- uv Python package manager for vector search, auto-installed if missing.
- Claude Code latest version with plugin support, or another supported harness.

### First steps

1. Run the installer, then sign in through the browser or pass an explicit --provider flag to skip sign-in.
2. Choose a memory provider: the claude-mem observer, your own OpenRouter or Gemini key, or your Anthropic plan.
3. Restart Claude Code. Context from previous sessions will appear in new sessions automatically.
4. Open the worker URL printed on startup to view My Work, or /viewer for memory, sessions and settings.

## Who Claude-Mem is for

- Claude Code users who want project knowledge to carry over between sessions.
- Developers using several agents such as Codex, OpenCode or Gemini who want shared memory.
- Teams wanting shared plans and decisions through the optional TeamBrain workspace.

### Limitations

- Claude-Mem's npm install -g installs only the SDK/library and does not register plugin hooks or the worker service.
- Claude-Mem's hosted CMEM Pro observer costs $30/month after a 30 day trial; memory otherwise falls back to your Anthropic plan.
- Claude-Mem recalls synced notes in ChatGPT, but ChatGPT conversations are not automatically captured.
- Claude-Mem's effect on token use depends on your workload and provider.

## Usage data (the week of Oct 25–31, 2026)

- Board rank: #31 overall, #18 in Coding agents, momentum score 82
- Downloads that week: 34.4k (+73% vs 4-week average)
- New GitHub stars per week (last 12 weeks): Aug 9: 794; Aug 16: 766; Aug 23: 1.1k; Aug 30: 785; Sep 6: 555; Sep 13: 605; Sep 20: 539; Sep 27: 1.1k; Oct 4: 3.8k; Oct 11: 2.2k; Oct 18: 1.5k; Oct 25: 5.4k

## Claude-Mem alternatives

| Rank | Tool | Stars | Downloads/wk |
|---|---|---|---|
| #1 | [Cindy](https://index.devtoolwatch.com/tools/makecindy-cindy/) | 8.8k | – |
| #2 | [OpenMausBot](https://index.devtoolwatch.com/tools/milind-soni-openmausbot/) | 9.6k | – |
| #3 | [Orca](https://index.devtoolwatch.com/tools/stablyai-orca/) | 111.7k | – |
| #4 | [OpenBot](https://index.devtoolwatch.com/tools/copilotkit-openbot/) | 9.1k | – |
| #5 | [Agent Reach](https://index.devtoolwatch.com/tools/panniantong-agent-reach/) | 107.2k | – |
| #6 | [Skill Recorder](https://index.devtoolwatch.com/tools/microsoft-skill-recorder/) | 5.6k | – |

## Claude-Mem FAQ

### Is Claude-Mem free and open source?

Yes. Claude-Mem's source code is on GitHub (thedotmack/claude-mem) under the Apache 2.0 license, so you can use it for free. Its website also links to a pricing page (https://claude-mem.ai/pricing) for any paid plans.

### How do I install Claude-Mem?

Run `npx claude-mem install`. It needs Node.js >=20.12.0. The install section above has the other options from the project's README.

### Is Claude-Mem still maintained?

Yes. The thedotmack/claude-mem repository was last updated on Oct 21, 2026, and the latest release, v15.0.1 on GitHub, came out on Oct 20, 2026.

### How popular is Claude-Mem?

Claude-Mem has 107.8k GitHub stars as of Nov 1, 2026. It gained 5.4k new stars in the week of Oct 25–31, 2026, against a 12-week average of 1.6k a week. Its main package had about 34.4k downloads that week, +73% against its own 4-week average. That made it #18 among AI coding agents on the Devtoolwatch Board.

### Claude-Mem vs Graphify: how do they compare?

Claude-Mem and Graphify are both tracked by Devtoolwatch. Claude-Mem has 107.8k GitHub stars and Graphify has 140k. In the week of Oct 25–31, 2026, Claude-Mem's main package had 34.4k downloads and Graphify's had 480.7k. Graphify is an AI coding assistant skill that turns a codebase, with its docs, PDFs, images and videos, into a queryable knowledge graph. [Graphify review](https://index.devtoolwatch.com/tools/graphify-labs-graphify/)

### Claude-Mem vs MemPalace: how do they compare?

Claude-Mem and MemPalace are both tracked by Devtoolwatch. Claude-Mem has 107.8k GitHub stars and MemPalace has 60.2k. In the week of Oct 25–31, 2026, Claude-Mem's main package had 34.4k downloads and MemPalace's had 15.1k. [MemPalace review](https://index.devtoolwatch.com/tools/mempalace-mempalace/)

### What are the best alternatives to Claude-Mem?

Other AI coding agents on the Devtoolwatch Board in the week of Oct 25–31, 2026, by momentum: Cindy (#1), OpenMausBot (#2), Orca (#3), OpenBot (#4), Agent Reach (#5). The alternatives table above has their numbers.

### How does Claude-Mem work?

Claude-Mem uses lifecycle hooks to capture tool usage during sessions, compresses it into summaries with AI, stores them in SQLite, and injects relevant context into later sessions. A local Bun-managed worker serves search and a web viewer.

### Does Claude-Mem increase token usage?

Claude-Mem's layered search returns a compact index first and fetches full details only for selected IDs, which the README says gives about 10x token savings. Hosted observation moves memory generation off your agent's subscription. The homepage says the effect depends on workload and provider.

### Does Claude-Mem work with Codex?

Claude-Mem works with Codex. The installer for T3 Code supports Codex and Claude Code providers, and the repo lists Codex among supported agents.

### Is Claude-Mem safe for sensitive code?

Claude-Mem lets you exclude sensitive content from storage with privacy tags. Local storage uses a SQLite database, and cloud sync to cmem.ai is an additional layer.

## Sources

1. [thedotmack/claude-mem on GitHub](https://github.com/thedotmack/claude-mem)
2. [Claude-Mem README](https://github.com/thedotmack/claude-mem/blob/main/README.md)
3. [claude-mem on npm](https://www.npmjs.com/package/claude-mem)
4. [Claude-Mem website](https://claude-mem.ai)
5. [Release v15.0.1 (Oct 20, 2026)](https://github.com/thedotmack/claude-mem/releases/tag/v15.0.1)
6. [Release v15.0.0 (Oct 20, 2026)](https://github.com/thedotmack/claude-mem/releases/tag/v15.0.0)
7. [Release v14.0.1 (Oct 17, 2026)](https://github.com/thedotmack/claude-mem/releases/tag/v14.0.1)
8. [Devtoolwatch Board, the week of Oct 25–31, 2026](https://index.devtoolwatch.com/)