Agent Cowork Memory
io.github.rxthlessbeatsPublished by GitHub account rxthlessbeatsv0.3.2Updated Oct 29, 2026
Shared memory and delegation for Codex, Cursor, Claude Code, and OpenCode.
The publisher's own server: Memory.fit
Overview
Gives Codex, Claude Code, Cursor and OpenCode one shared project memory so an assistant can continue another agent's chat, share notes and delegate work.
- What you can do
- Read thread context, timelines, project notes, open threads, recent jobs and held files
- Save notes every agent sees, search notes and delegated job results, flag stale notes
- List recent chats across agents and resume a chosen chat into its thread
- Delegate work to other agents with a brief and wait for their results
- Run setup, status, doctor and backup commands from the CLI
- When to use it
- Useful when you switch between several coding agents on the same project and want context to carry over instead of pasting summaries, or when you want one agent to hand tasks to another in the background.
- Requirements
- Local process run with npx; Node 24 or newer. Setup adds it to Cursor and to Codex, Claude Code and OpenCode when installed, then agents must be restarted. State lives in a folder under the home directory, movable with ACM_HOME. Optional herdr for visible panes.
Installation
In SourceWeft
- Open Agent Cowork Memory in the dashboard and add it to a workspace.
- Enable the server for the chats that should use its tools.
Desktop only via STDIO. STDIO servers start a local process, so they need the SourceWeft desktop host.
Other MCP clients
Follow the launch instructions in the repository.
README
Agent Cowork Memory (ACM)
Hit your Codex limit mid-task? Open Cursor, say "continue from Codex", and keep going. ACM gives Codex, Claude Code, Cursor, and OpenCode one shared memory, so you never paste context again, and any agent can put the others to work.
Verified by benchmark: with Codex on 38 held-out tasks, ACM finished 223 of 228 runs, against 130 with a pasted summary. Cursor and OpenCode performed similarly with ACM and summaries.
[GitHub Release]
[npm]
[CI]
[License]
[M8ven Score]
[Agents supported]
[Node]
[Benchmark]
[MCP supported]
[CLI supported]
[Codex icon] Codex · [Claude Code icon] Claude Code · [Cursor icon] Cursor · [OpenCode icon] OpenCode
Demo
[Claude Code delegates to Codex, Cursor, and OpenCode through ACM]
More demos on the ACM official site
Install
With Node 24 or newer:
Setup adds ACM to Cursor, and to Codex, Claude Code, and OpenCode when they are installed; Installed one later? Run setup again. Restart your agents, and ACM is available in every project.
Tested for real on Linux and WSL. macOS and Windows pass CI; real runs welcome.
Use
Memory Sharing
Open the same project in Codex and Cursor. To pick up a chat in the other agent, ask:
Use ACM to continue from Cursor.
Or, in Cursor:
Use ACM to continue from Codex.
Every agent can see every chat in the project. If there is more than one, the agent lists them (agent, time, thread, first message, and whether ACM started it for a delegated job) and asks you which one to continue. You can also name it:
Use ACM to continue the Codex chat about the trip.
Related work goes together in a thread: continuing a chat, or delegating from it, keeps the work on one thread, and a new chat can join one of the project's open threads. ACM keeps each thread's timeline as things happen (chats joining, jobs starting and finishing, notes added), so any agent can see where the work stands.
Ask your agent to remember a decision, and every agent in the project sees it. Notes stay true.
Delegate to other agents
Any agent can hand work to the others:
Tell codex to write plan.md, and claude to write packing.md.
Each agent gets a brief, joins the thread it was sent from, claims the files it edits so they don't collide, and reports back. Agents run in the background. With herdr installed, each one gets its own pane you can watch (herdr session attach acm). If you resumed a chat that's open in a herdr pane, work for that agent goes back to that chat, in its pane.
Benchmark
What it evaluates
- How much of earlier chats' context survives when a new chat continues the work.
- Each task is a small Python repo in two steps: earlier chats do step 1 and leave step 2's details in the chat.
- The details sit in one short or long chat, spread over two chats, changed later, corrected mid-chat, or against a README that still says the old thing.
- A run counts only if every hidden check passes, existing tests still pass, and no old or replaced value is implemented.
What we tested
- Codex (gpt-6-luna, medium effort), Cursor (Grok 4.7 High), and OpenCode (GLM-5.3-Flash), continuing the work five ways:
- no ACM;
- ACM ("Use ACM to continue from my last chat…");
- the summary the earlier chat writes when asked;
- a careful handoff note, including any Markdown files its turn wrote;
- every earlier chat pasted in.
- Held-out tasks written before the runs, with the analysis fixed in advance. Pilot checks and earlier-chat guards determine which tasks count for each agent.
- Two independent histories, 3 runs per way per history; no ACM runs once per history.
MCP tools
ACM tells which chat it is in on its own; no tool needs a session or, from the project folder, a path (every tool takes an optional repo_path).
CLI
--home <dir> (or ACM_HOME) moves ACM's state folder from ~/.agent-cowork-memory. Run them with npx -y agent-cowork-memory <command>, or acm once it's installed globally.
Source: README.md at commit be804bc
Tools
0Version history
4- v0.3.2LatestOct 29, 2026
- v0.3.1Oct 20, 2026
- v0.3.0Oct 6, 2026
- v0.2.1Sep 30, 2026
Score
50% of the applicable checks met
How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.
- Reached in the last 30 days25 pts
Keep the endpoint answering the MCP handshake; SourceWeft probes it every 30 days.
- Auth declared10 pts
Declare the credential headers in the registry entry, or keep the endpoint answering the probe so its auth can be classified.
- Tools listed15 pts
Answer tools/list without a credential, or ship a package that starts in a sandbox without one.
- DescriptionRequired15 pts
- Logo10 pts
- Documentation15 pts
- License10 pts
Other Memory integrations
Memory.fit
fit.memoryOfficial publisher
Hosted Memory.fit server that lets an AI read your training, sleep, recovery and nutrition data and write back weekly plans, targets and meal plans.
pg.ddx.io: the PostgreSQL community's memory
io.ddx.pgOfficial publisher
pg.ddx.io: the PostgreSQL community's memory's hosted MCP server, published by pg.ddx.io; no sign-in required; Streamable HTTP; 29 tools.
var.gg Naming Memory
gg.varOfficial publisher
Remote MCP server that suggests and searches code identifier names from domain concepts in Korean, Japanese, Chinese, Vietnamese and English, and exposes ERD…
Memoria — Deterministic AI Memory & Code Intelligence MCP
Published by GitHub account kolivritech
Remote MCP server giving AI coding agents deterministic long-term memory plus code intelligence tools, with no LLM tokens spent on recall.
thron-memory
Published by GitHub account jacobbarglof
Gives every MCP-capable AI app one shared, private memory of plain note pages that assistants can search, read and update.