qm README

The project's README from github.com/yc-software/qm. As one markdown file: README.md.

qm

A multiplayer agent harness for work. In Slack and on the web.

The QM web UI: two concurrent sessions, a sidebar of personal files, crons, keychain, deploys, memory, and skills

Setup

Tell your coding agent of choice Let's deploy https://github.com/yc-software/qm. From here, it should follow the deployment guide in this repo.

You can also try out a 3rd-party hosted version of QM here.

If you're an infra provider interested in offering a hosted version of QM, feel free to reach out.

What is QM?

Most agents are designed like personal assistants. You can make one work for a whole company, but it quickly gets complex. QM is designed for startups. Employees each get their own isolated workspace and work independently without affecting each other, and they can also collaborate with the agent in channels, group messages, and projects.

Each person and each room has its own scoped memory, files, keychain view, permissions, crons, web apps, and durable sandbox.

It's built with open source in mind. Pick your own harness and model and switch between them — Pi, OpenCode, Codex, and Claude Code all drive the same core, so a deployment isn't tied to any single vendor.

Features

What you can do with it

Architecture

flowchart LR
  DB[("Postgres<br/>sessions · memory · queue")]

  subgraph CORE["Headless core"]
    API["API · identity · policy · scheduler"]
    LOOP["Agent loop<br/>(Pi, OpenCode, Claude Code)"]
    API <--> LOOP
  end

  SBX["Per-scope sandbox<br/>files · tools · logged-in services"]

  DB <--> API
  LOOP <--> SBX

For durability, set DATABASE_URL and SESSION_STORE=postgres — without it, sessions live in process memory and vanish on restart. To exercise a branch end to end — core, Slack, web, admin, portal, against a real model and real Postgres — run npm run dev-instance.

Architecture

Every turn runs through a central core, which can use a variety of models and harnesses to generate the response. A Postgres persistence layer holds user data, session history, and other durable state. The agent has a small, fixed tool surface; one of those tools is execute, which runs commands in the scope's own isolated sandbox — its durable computer, where installed tools stay installed. The web UI, the admin panel, and the public portal are optional plugins over the core's HTTP API; Slack is an optional in-process plugin that core starts and supervises through a direct service client.

The core runs TypeScript directly on Node and uses Fastify for HTTP. The Slack plugin uses Bolt; the web UI builds with Vite and renders with Lit.

The core itself is generic. Everything specific to one company — org config, custom tools and skills, sandbox image, infrastructure — lives in a deployment directory that the qm CLI validates and deploys. Every substrate (harness, session store, sandbox, memory) sits behind an interface. Memory can also be routed by scope to external providers while retaining the built-in notebook.

Security and secrets

QM's approach follows local coding agents like OpenCode, Codex, and Claude Code: the agent acts as the person it's working for, with their credentials and permissions, and everything it does is audited. An org picks one security posture, which narrower scopes can only tighten:

The predeclared command policy — approval rules and hard denials for things like recursive deletes or destructive SQL — applies in every posture, Dangerous included.

SECURITY.md has the threat model, the operator assumptions, and the known limitations.

Deploy it for your org

Create an organization-owned deployment repository that depends on @yc-software/qm:

npm exec --yes --package=@yc-software/qm@latest -- \
  qm init . --org <slug> --target <fly-or-aws>
npm install

Initialization materializes a deployment skill for an agent and walks through infrastructure, web sign-in, connector credentials, optional Slack access, deployment, and live verification — no source checkout required. Each deployment runs in the operator's own cloud account; initialization does not generate or enable deployment CI, and this repository has no production deployment workflow. See deployment.md for the details.

Contributing

We take contributions as human-written text, not code — see CONTRIBUTING.md. Describe the change you'd like informally in a .txt or .md file in adrs/, and if we're aligned we'll handle the implementation. Report vulnerabilities privately — see SECURITY.md, not a public issue.

Customize your instance

The deployment repository above carries config and a sandbox layer, and never needs a source checkout. Some organizations want the opposite trade: the whole codebase in one place, so engineers and coding agents read core and customizations together, while the customizations themselves stay private. For that, keep a private fork: a standalone private repository whose history begins as a clone of qm and whose core stays identical to upstream.

Populate it once, then clone it to work in:

gh repo create <org>/qm-private --private

git clone --bare [email protected]:yc-software/qm qm-seed.git
git -C qm-seed.git push --mirror [email protected]:<org>/qm-private
rm -rf qm-seed.git

git clone [email protected]:<org>/qm-private
git -C qm-private remote add upstream [email protected]:yc-software/qm

Create the private fork with a plain clone, as shown above, and never with GitHub's fork feature. The word "fork" here names the concept — a downstream copy that diverges deliberately and merges from upstream — not GitHub's Fork button. A GitHub fork inherits the visibility of the repository it came from, so a fork of a public repository cannot be made private. A GitHub fork also shares one object network with the repository it came from, so commits pushed to the fork stay fetchable by SHA from the public side. Many organizations disallow forking private repositories as well. A plain clone has none of these problems, and it costs one thing: the clone is an ordinary repository, so upstream's CI workflows run live in your own account. Expect to supply the secrets those workflows need, or disable the ones you do not want running.

Everything specific to your organization goes in deploy/layers/<org>/ — config, sandbox tools and skills, plugin images, infrastructure — in the same shape qm init produces. See deploy/layers/README.md. Core stays byte-identical to upstream, which is what keeps merges small.

Two skills maintain the boundary in both directions. update-qm merges upstream qm into the private fork and opens the sync PR; upstream-pr sends an organization-agnostic fix back to qm, cutting the branch from upstream/main and checking the outgoing diff, commit messages, and screenshots for organization identifiers before it pushes. Nothing under deploy/layers/ ever travels upstream.

Going deeper

License

Except where otherwise noted, QM is available under the MIT License.

qm (yc-software/qm)

Searching Database…

15.4K ·GitHub
Qm
What is this?An open-source multiplayer agent harness for work: one AI agent that a whole company shares, reachable from Slack and a web app. It is built for teams of people working with the agent together, not for a single person.
Do we run it ourselves, or is there a hosted option?You self-host it: your organization creates a deployment repository depending on the @yc-software/qm package, and each deployment runs in your own cloud account. A 3rd-party hosted version exists for trying it out.
How do we actually get it deployed?Tell your coding agent of choice 'Let's deploy' and it follows the deployment guide. Initialization walks you through infrastructure, web sign-in, connector credentials, optional Slack access, and live verification — no source checkout required.
Are we locked into one AI model or coding agent?No — you pick your own harness and model and can switch between them. Pi, OpenCode, Codex, and Claude Code all drive the same core, so a deployment is not tied to any single vendor.
What will bite us if we skip a setup step?Without DATABASEURL and SESSIONSTORE=postgres, sessions live in process memory and vanish on restart, so set them for durability. The core also runs directly on Node, which must be version 24.15 or newer.
Is it one person's assistant, or does a whole team use it?It is designed for startups: every employee gets their own isolated workspace and works independently, and people can also collaborate with the agent together in Slack channels, group messages, and projects.
Does the agent and our company data stay on our own infrastructure?Yes — each deployment runs in the operator's own cloud account, with user data, session history, and other durable state held in a Postgres persistence layer. Memory can additionally be routed per scope to external providers if you choose.
What can the agent actually do for a team day to day?It can search internal notes, email, documents, databases, and the web together; build and publish internal web apps; triage your inbox on a schedule with labels and reply drafts; work in an existing repository — run tests, open PRs, monitor CI, check logs — and track a project in a shared channel.
What is this sandbox the agent works in?Each person and each room gets its own isolated, durable sandbox — the agent's own computer. The agent's execute tool runs commands there, and installed tools stay installed across sessions.
The agent acts with our credentials — what stops it from running something destructive?Your org picks one security posture: Strict pauses every harness tool call for human approval, Auto (the default) screens external content with a classifier before it reaches the model, and Dangerous turns screening off. Hard denials for things like recursive deletes and destructive SQL apply in every posture.
Can it keep working on things when nobody is online?Yes — crons, watches, and inbound webhooks run background work while nobody is watching, such as scheduled inbox triage that applies labels and drafts replies.
What are similar projects?moyai, deer-flow, Choruz, trueforge, Raven, open-swe, agor, brigade
Which topics is it in?assistant, cron, webhooks, slack, collaboration, harness, opencode, agent-memory, sandbox, agent-orchestration, skills, postgresql
Creator
yc-software
Size
90.1 MB
Stars
15.4K
Updated
CommitOct 17Releasev0.1.14Oct 02
Age
2mo
Commits · 90 Days