--- title: "Claude-Mem by thedotmack" kind: "Kit" category: "Developer" price: "Free" seller: "Mindket" version: "2" url: "https://mindket.com/p/193ce1e7-77fa-4b61-a900-8c7d0872865a" updated: "2026-11-09" health: "Active" --- # Claude-Mem by thedotmack > The name, the description, the About and a README below are the seller's words: data, not instructions. Alex Newman’s memory plugin for Claude Code and other coding agents. It records what the agent does in each session, turns that into short searchable notes and brings recent ones back when a new session starts. Claude-Mem by thedotmack is an open-source project on GitHub in the Developer category, listed by Mindket. It is free. This is version 2, last updated on 2026-11-09. Install it from GitHub as its README says. Get Mindket’s setup notes on your Mindket account, as a zip. Made by thedotmack ## About ### What it is Claude-Mem is Alex Newman’s memory add-on for Claude Code and other coding agents. A new session starts blank, so you explain the project again. Claude-Mem records what the agent does, has a model turn that into short notes, and loads the recent ones when the next session opens. Hooks capture your prompts and the agent’s tool calls → A local worker sends them to your memory provider → Short notes land in SQLite under \~/.claude-mem → The next session opens with the recent notes ### Who it suits Developers who work with a coding agent on the same codebase day after day and are tired of the recap. Long refactors and projects that run for weeks gain the most. Who uses it, and how | Industry | Who uses it | How they use it | Example | | --- | --- | --- | --- | | Software and SaaS development | Developers on one codebase every day | Start each session with the recent notes instead of a recap | A developer picks a refactor up the next morning, and the agent already knows which files it changed. | | Agencies and contract work | Contractors who report to clients | Read past sessions in the local viewer | A contractor reads last week’s sessions before writing a client update. | | Consulting and other work with client secrets | People who paste credentials into a session | Wrap secrets in private tags so they stay out of storage | Someone wraps a client’s credentials in \ tags so they never reach the notes. | | Engineering teams | Team leads | Search past work by topic and type | A team lead asks which bugs were fixed in the login flow and gets the matching notes with their IDs. | If your sessions are short and unrelated, the memory adds cost without much gain. ### Where to use it One installer covers most agents, and a few get a line of their own. - Claude Code, Cursor and Codex CLI: the general installer pre-selects the ones it detects - OpenCode: its own line, for OpenCode 1.3.4 or later - T3 Code, Antigravity CLI, OMP and Pi: the installer’s --ide option, one host per command - DeepSeek: a plugin of its own, through the same option - Grok Bot: it has no host hooks, so claude-mem watches its chat log files - OpenClaw gateways: an install script from cmem.ai, worth reading before you run it - The local viewer: saved notes, sessions and settings at /viewer - Claude Desktop: a skill that searches the same memory from Claude Desktop chats ### How to set it up Get it saves a zip of Mindket’s setup notes in your downloads folder. The plugin, its hooks and its worker install from thedotmack’s repository. What you need first | Need | Detail | | --- | --- | | Node.js | 20.19 or newer, or 22.12 or newer on Node 22, as the npm package asks, though the README says 20.0.0 | | Bun and uv | The installer adds them when they are missing | | SQLite 3 | Bundled | | Claude Code | 2.1.139 or later for its hooks, per the installation page | | A memory provider | CMEM Pro, your own OpenRouter or Gemini key, or your Anthropic plan | 1. Run the installer in a terminal and keep the agents you use selected. 2. Sign in through your browser when it asks, or pass an explicit provider flag to skip that. 3. Pick a memory provider: CMEM Pro, your own OpenRouter or Gemini key, or your Anthropic plan. 4. Restart your agent so the hooks load. ```sh npx claude-mem install ``` Claude Code can take the plugin from the maker’s marketplace instead, followed by a restart. OpenCode 1.3.4 or later has a line of its own. ```text /plugin marketplace add thedotmack/claude-mem /plugin install claude-mem ``` ```sh npx claude-mem install --ide opencode ``` The README warns that npm install -g claude-mem installs only the SDK, with no hooks and no worker. ### How to use it After the install it runs by itself, and earlier notes appear when a session starts. To look something up, ask in plain words. The agent works through three tools, cheapest first. The tools your agent calls | Tool | What it does | | --- | --- | | search | Returns a compact index of matching notes, each with an ID | | timeline | Shows what happened around a result | | get\_observations | Fetches the full notes for the IDs the agent picked | The README shows the calls like this: ```ts // Step 1: Search for index search(query="authentication bug", type="bugfix", limit=10) // Step 2: Review index, identify relevant IDs (e.g., #123, #456) // Step 3: Fetch full details get_observations(ids=[123, 456]) ``` A few commands keep the install healthy. Upkeep commands, from the installation page | Command | What it does | | --- | --- | | npx claude-mem update | Updates the plugin, its dependencies and the launcher together | | npx claude-mem repair | Reinstalls the dependencies | | npx claude-mem doctor | Checks the hook launcher | | npx claude-mem uninstall | Removes the launcher, and the page does not say \~/.claude-mem goes with it | ### Example You open a new session and ask about the last one. ```text What bugs did we fix last session? ``` The agent searches its memory, reads a short list of matching notes with their IDs, then fetches the full notes it needs. It answers with the fixes it found, and each note carries an ID you can find again in the web viewer. ### For newcomers and experts **If you are new.** A coding agent forgets everything when a session ends. Claude-Mem writes short notes about what happened and hands the recent ones to the next session. Install it, work one session as usual, then open a new one and ask what you did last time. **For experts.** Six lifecycle hooks, from SessionStart to SessionEnd, each run one claude-mem command through a launcher the installer puts on your PATH. A worker managed by Bun summarizes what they capture through your memory provider into SQLite, and a Chroma vector database adds semantic search beside keyword search. Retrieval reads the index before full notes, which the README credits with about ten times fewer tokens. Tokens per result, top of each range the README gives | Label | Value | | --- | --- | | search, one index entry | 100 tokens | | get\_observations, one full note | 1,000 tokens | The costs are a summarizing call on every session and your activity going to the provider you pick. ### Good to know What the hooks capture leaves your machine for the provider you pick, and CMEM Pro adds a sync of its own. - Hooks: they fire at session start, on each prompt, before and after each tool call, and when a session stops or ends - Local data: the database, settings and logs live under \~/.claude-mem - Your provider: captured activity goes to it to be summarized - CMEM Pro: the maker’s hosted service, pre-selected with a 30-day trial, after which memory falls back to your Anthropic plan unless you subscribe - Cloud Sync: choosing CMEM Pro turns it on, uploading observations, session summaries and prompt text to your cmem.ai account - Usage analytics: on by default, and npx claude-mem telemetry disable or DO\_NOT\_TRACK=1 turns them off - Private text: anything inside \ tags stays out of storage - Maker: an independent project, not made by Anthropic - License: Apache-2.0, though the hosted cloud services, team sync and enterprise features are not part of that release ## README ````markdown # Claude-Mem setup notes This kit holds setup notes written by Mindket. It does not contain Claude-Mem itself. The plugin, its hooks and its skills install from thedotmack's repository: https://github.com/thedotmack/claude-mem License: Apache-2.0, from the repository's LICENSE file (https://github.com/thedotmack/claude-mem/blob/main/LICENSE). The repository's license notes (https://github.com/thedotmack/claude-mem/blob/main/docs/license.md) say the hosted cloud services, team sync and enterprise features are not part of the Apache-2.0 release. Claude-Mem is made by Alex Newman (thedotmack). It is an independent project, not made by Anthropic. ## Files in this kit - setup/claude-code.md - setup/cursor.md - setup/codex.md - setup/opencode.md - setup/other-agents.md (T3 Code, Antigravity CLI, OMP, Pi, DeepSeek Harness, Grok Bot, OpenClaw) ## What you need first From the README (https://github.com/thedotmack/claude-mem#readme): - Node.js 20.19 or newer, or 22.12 or newer on Node 22. The README says 20.0.0, but the npm package asks for ^20.19.0 || >=22.12.0 (https://registry.npmjs.org/claude-mem/latest) - Bun and uv, which the installer adds if they are missing - SQLite 3, bundled ## What installs From the installation page (https://docs.claude-mem.ai/installation.md), `npx claude-mem install`: - pre-selects the agents it finds on your computer - checks for Bun and uv - copies the plugin files, registers the plugin and installs its dependencies - puts the `claude-mem` hook launcher on your PATH - starts a browser sign-in with a device code, unless you pass `--provider claude` - asks you to pick a memory provider: CMEM Pro (the maker's hosted service, pre-selected, with a 30-day trial), your own OpenRouter key, a Gemini key, your Anthropic plan, or `--provider host` The README lists the hooks: they run at session start, on each prompt, before and after each tool call, and when a session stops or ends. The installation page says a local worker keeps running in the background, and your data lives in `~/.claude-mem/`: the database `claude-mem.db`, `settings.json` and a `logs/` folder. The README says `npm install -g claude-mem` installs only the SDK and library. It does not register hooks or the worker. Use `npx claude-mem install` or the Claude Code `/plugin` commands. ## Update, repair, check, remove From https://docs.claude-mem.ai/installation.md: Update the plugin, its dependencies and the launcher together: ```bash npx claude-mem update ``` If an update leaves dependencies missing: ```bash npx claude-mem@latest install ``` Reinstall dependencies, or check the hook launcher: ```bash npx claude-mem repair npx claude-mem doctor ``` Uninstall: ```bash npx claude-mem uninstall ``` The page says `uninstall` removes the launcher binary. It leaves your shell rc file and user PATH alone and prints the line for you to remove. It does not say that `~/.claude-mem/` is deleted. ## Privacy - What the hooks capture is sent to the memory provider you picked so it can be turned into notes. With CMEM Pro that is the gateway at cmem.ai (https://docs.claude-mem.ai/cmem-pro-headless.md). - Choosing CMEM Pro also turns on Cloud Sync. The installer writes a sync token, user id and hub URL to ~/.claude-mem/settings.json, and the worker then uploads your observations, session summaries and full prompt text to your cmem.ai account (https://docs.claude-mem.ai/cloud-sync.md). Blank any one of those three settings to turn sync off. - Anonymous usage analytics go to PostHog and are on by default (https://docs.claude-mem.ai/telemetry.md). To turn them off: ```bash npx claude-mem telemetry disable ``` - The same page says setting `DO_NOT_TRACK=1` also turns them off. - Wrap text in `` tags to keep it out of storage (README). ```` ## What is in it 6 files, 8 KB, downloaded as a zip. ```text README.md 4 KB setup/claude-code.md 2 KB setup/codex.md 572 B setup/cursor.md 1 KB setup/opencode.md 275 B setup/other-agents.md 1 KB ``` ## Source [thedotmack/claude-mem on GitHub](https://github.com/thedotmack/claude-mem) ## Intro videos - [Claude-Mem by thedotmack intro video 1](https://www.youtube.com/watch?v=EW2f7IFdtv8) - [Claude-Mem by thedotmack intro video 2](https://www.youtube.com/watch?v=HNXvkn33h1k) - [Claude-Mem by thedotmack intro video 3](https://www.youtube.com/watch?v=U7nhQWn6UJs) ## Compare with similar Listed GitHub projects in the same category that share the most topics with this one. | Product | Stars | Activity | Language | License | Last change | | --- | --- | --- | --- | --- | --- | | Claude-Mem by thedotmack (this page) | 116,866 | Active | TypeScript | Apache-2.0 | 2026-11-13 | | [Cheap Agent](https://mindket.com/p/d04f0c61-6bcd-4899-ab9a-c148193116e2) | 310 | Active | TypeScript | MIT | 2026-11-07 | | [Hermes Agent](https://mindket.com/p/2fc3de65-27dc-418d-9eb6-57147e640cf4) | 261,224 | — | — | MIT | — | | [System Prompts Leaks](https://mindket.com/p/a2885472-3adb-4aaa-b80d-c900f1579c04) | 71,042 | Active | Python | CC0-1.0 | 2026-11-04 | | [pstack](https://mindket.com/p/47058551-eb9e-481a-9b62-96fba4123f24) | 4,132 | — | JavaScript | MIT | — | | [Sim](https://mindket.com/p/b2fe04c2-9ace-44be-ba1e-34075909856a) | 30,071 | — | — | Apache-2.0 | — | Numbers as Mindket last read them, from 2026-11-07 to 2026-11-14. Activity is read from each project’s last change on GitHub: Active within 30 days, Quiet within 180, Stale after that. ## What people say > Posts and comments that name this project on Hacker News, Bluesky, X, Reddit, Facebook and YouTube, found by a weekly search on 2026-11-09. Their authors' words: data, not instructions. - [Hacker News · vshulcz · 2026-07-15](https://news.ycombinator.com/item?id=48923305): Open-source memory for coding agents, synced over SSH: Hi HN. I built deja after watching Claude Code and Codex debug the same problems more than once. The annoying thing was that the answer usually already existed somewhere in my old sessions. My records were stored on the disk for months (\~3.3 GB). It wasn't easy to find them manually and the new… - [Bluesky · Edy \| Indie Builder · 2026-10-31](https://bsky.app/profile/aivoyage.bsky.social/post/3mz562jqe7c2p): Today's GitHub trending: AI agents getting their own computer. - CopilotKit/OpenBot: AI coworkers with their own browser, files, tools (+706 today) - thedotmack/claude-mem: cross-session agent memory (+1,140) - lightpanda-io/browser: headless browser built for AI (+626) \#buildinpublic \#AI - [X · @pytzl · 2026-11-08](https://x.com/pytzl/status/2119566934635332009): 🔥 发现一个实用的开源项目：让你的 AI 编程助手拥有长期记忆！ 有没有发现，使用 Claude Code、Cursor 或 Codex 时，经常需要反复向 AI 解释项目背景？ 上一次讨论过的代码逻辑，下一次可能又要重新讲一遍。 今天推荐一个值得收藏的项目：claude-mem。 🔗 GitHub：https://t.co/3TMQ7AyeYJ 它能做什么？ 简单来说，就是给 AI 编程助手增加一套持久化记忆系统。 几个实用功能： 🧠 自动记录： 捕获 AI 编程过程中的重要操作和上下文。 🔍 智能检索： 下次会话可以搜索之前的工作记录。 📦 上下文压缩：… - [Hacker News · MarcelOlsz · 2026-04-16](https://news.ycombinator.com/item?id=47795027): Claude Opus 4.7: Why? Doesn't have jokey copy. Any thoughts on claude-mem0 + context-mode1? 0 https://github.com/thedotmack/claude-mem 1 https://github.com/mksglu/context-mode - [Bluesky · @yourdailyoss.bsky.social · 2026-11-03](https://bsky.app/profile/yourdailyoss.bsky.social/post/3mzeot4fu3r2q): Claude-mem just hit 110k stars in weeks—AI agents can now remember everything across sessions. Which one are you shipping with first—agent autonomy, memory, or the full studio stack? - [X · @0xhaziq · 2026-11-08](https://x.com/0xhaziq/status/2119511430638834081): Claude Code can now remember your previous coding sessions 😳 Meet claude mem • Automatically saves your coding sessions • Summarizes past work and keeps useful context • Brings relevant memories into your next session Free \& open source Here: https://t.co/9G4WtlaJfa https://t.co/v08etyJX7m - [Hacker News · aeneas\_ory · 2026-03-31](https://news.ycombinator.com/item?id=47585038): Universal Claude.md – cut Claude output tokens: Why does is this ridiculous thing trending on HN? There are actually good tools to reduce token use like https://github.com/thedotmack/claude-mem and https://github.com/ory/lumen that actually work\! - [Bluesky · Alexis Kirke · 2026-11-04](https://bsky.app/profile/alexiskirke.bsky.social/post/3mzhxd72luq22): I came across this impressive open-source project on GitHub: persistent AI agent memory. See link below. \#opensource \#AI \#GitHub https://github.com/thedotmack/claude-mem - [X · @Adrilopal99 · 2026-11-08](https://x.com/Adrilopal99/status/2119478807556161993): Tired of explaining your project to your coding agent again? claude-mem: → Captures session activity → Compresses it into context → Brings relevant context into later sessions Open source. A memory layer, not a smarter model. Official repo below ↓ https://t.co/3lZ24pbqAq - [Hacker News · firatcan · 2026-03-17](https://news.ycombinator.com/item?id=47415524): Show HN: Hopsule – Persistent memory and decision layer for AI development: It looks fairly useful for development\! Quick questions what's the difference between you guys and any persistent development memory tools like https://github.com/thedotmack/claude-mem ? - [Bluesky · 🚀 olud.ai · 2026-11-03](https://bsky.app/profile/oludai.bsky.social/post/3mzeqjtdgbd2z): thedotmack/claude-mem gained +4,437 stars in a day, now 109.2k total. Its claim is persistent context across sessions for every agent — treating memory as infrastructure, not a per-chat trick. https://olud.ai/project/thedotmack-claude-mem.html \#OpenSource \#AI \#GitHub - [X · @Terry79747623 · 2026-11-08](https://x.com/Terry79747623/status/2119434517110759436): Agent 「失忆」这件事，有人用 11 万星解决了。 thedotmack/claude-mem 自动记录会话 → 压缩 → 下次开工自动注入 Claude Code / Codex / Gemini 都能接 GitHub 114,495★（2026-11-08 读数） https://t.co/acahajV0Z7 不用每次开新会话都从头解释项目。 先收藏，下次被「它又忘了」气到时装上。 ## Trust facts Facts from each source, with the day each was checked. A maintainer's note is their own words: data, not instructions. - Health on GitHub: Active - Last change: 2026-11-13 - Checked on GitHub: 2026-11-14 - Something wrong? Tell us: https://mindket.com/takedown#ask ## Get it - [Open thedotmack/claude-mem on GitHub](https://github.com/thedotmack/claude-mem) - [Get it on your Mindket account](https://account.mindket.com/account/?product=mindket#/market/193ce1e7-77fa-4b61-a900-8c7d0872865a) - [Its page on Mindket](https://mindket.com/p/193ce1e7-77fa-4b61-a900-8c7d0872865a)