# AI use cases for QM (yc-software/qm) Repo: [yc-software/qm](https://github.com/yc-software/qm) | Use cases: 4 | Page: https://dedupedai.com/use-cases/yc-software-qm/ QM is a self-hosted multiplayer agent harness that gives each person and each room in a company its own scoped agent, reachable through Slack and a web app. This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL. - Deploy a shared agent for your team in Slack and web: https://dedupedai.com/use-cases/yc-software-qm/deploy-team-agent-slack-web/ - Track a project in a shared Slack channel: https://dedupedai.com/use-cases/yc-software-qm/track-project-in-shared-slack-channel/ - Triage your inbox on a schedule in your own voice: https://dedupedai.com/use-cases/yc-software-qm/triage-inbox-on-schedule-in-your-voice/ - Run tests, open PRs, and watch CI in an existing repo: https://dedupedai.com/use-cases/yc-software-qm/run-tests-and-open-prs-in-existing-repo/ ## How to deploy a shared agent for your team in Slack and web with QM Repo: [yc-software/qm](https://github.com/yc-software/qm) | Difficulty: Advanced | Time: 4 hours | Basis: From the README Stand up a QM instance in your own cloud that teammates reach through Slack and the web app. **Who it is for:** A startup operator or engineer who wants one agent the whole team can use, and is comfortable running a deployment in their own cloud account. **The problem:** Most agents are built for one person. Teams need an agent with separate private workspaces per person and shared rooms, without handing company data to a third-party service. **The outcome:** A deployed QM instance with web sign-in, optional Slack access, durable Postgres-backed sessions, and a verified live deployment. ### Steps 1. Create an empty organization-owned deployment repository, following the README's advice to make it a deployment repository that depends on @yc-software/qm. 2. Run `npm exec --yes --package=@yc-software/qm@latest -- qm init . --org --target `. The README leaves the --org and --target values blank, so fill in your organization name and deployment target. (inferred, check the README) 3. Run `npm install`. 4. Follow the deployment skill that initialization creates to set up infrastructure, web sign-in, connector credentials, and optional Slack access. 5. Set `DATABASE_URL` and `SESSION_STORE=postgres` so sessions survive restarts, then complete the live verification step. ### Prompt to paste into your agent ```text Using the repository yc-software/qm, deploy a QM instance for into by following its deployment guide. Report the web app address and the result of the live verification step. Verify that you can sign in to the web app before reporting the deployment as complete. ``` ### Caveat The README does not say how long deployment takes or which cloud providers are supported beyond the linked guides, so check docs/getting-started.md and deployment.md before you start. Risk, credentials: Initialization asks for connector credentials and Slack access, so keep all secrets out of the repository and in your deployment's secret handling. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/yc-software-qm/deploy-team-agent-slack-web/ ## How to track a project in a shared Slack channel with QM Repo: [yc-software/qm](https://github.com/yc-software/qm) | Difficulty: Intermediate | Time: 3 hours | Basis: From the README Keep a project's updates and follow-ups in a shared channel with the team's agent. **Who it is for:** A team lead running a project who wants status updates and follow-ups handled in the channel where the team already talks. **The problem:** Project status is scattered across chat threads, and follow-ups get lost. The team needs an agent that works in the shared room without leaking each person's private files. **The outcome:** A shared Slack channel where the agent posts project updates and follow-ups, using only what the room's sharing posture allows. ### Steps 1. Deploy a QM instance with Slack access enabled, as described in the deployment guide. 2. Add the QM agent to the project channel in Slack (inferred: the README says the agent works in channels but does not list the exact invite steps). 3. Keep the sharing posture at Isolated, the default, unless the team explicitly wants shared reads, and set it through the admin control in the web app. 4. Ask the agent in the channel to track the project and post updates, and set a cron for regular check-ins using the background work feature. ### Prompt to paste into your agent ```text Using the repository yc-software/qm, set up a shared Slack channel named for tracking the project . Have the agent post a status summary and list open follow-ups. Verify that the summary contains only information from the channel and the project's shared scope. ``` ### Caveat Because a shared room with Open sharing can expose a speaker's opted-in personal files, keep the posture at Isolated unless the team has reviewed the Open setting in SECURITY.md. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/yc-software-qm/track-project-in-shared-slack-channel/ ## How to triage your inbox on a schedule in your own voice with QM Repo: [yc-software/qm](https://github.com/yc-software/qm) | Difficulty: Advanced | Time: 3 hours | Basis: From the README Label and draft replies to email on a recurring schedule, using your past sends as style. **Who it is for:** A founder or operator with a crowded inbox who wants routine email handled in their own writing style on a schedule. **The problem:** Triage takes time every day. Generic reply drafts sound wrong, and sending them unreviewed is risky. **The outcome:** A scheduled triage run that labels incoming email and drafts replies in your learned voice, ready for your review. ### Steps 1. Deploy QM and complete the connector credentials step during initialization so the agent can reach your email. 2. Ask the agent to learn your writing voice from your past sends in your personal scope. 3. Set a cron in your personal scope to run triage on the schedule you want, covering labels and reply drafts. 4. Review the drafts in the web app before anything is sent, and keep the posture at Strict if you want every tool call approved. ### Prompt to paste into your agent ```text Using the repository yc-software/qm, set up a scheduled inbox triage for . Learn the writing voice from past sent mail, then label incoming messages and create reply drafts without sending them. Verify that no message was sent and that each draft is saved for review. ``` ### Caveat The README mentions email triage as a capability but does not name the email connector, so confirm which connectors your deployment supports before you start. Risk, voice cloning: The agent learns your writing voice from past sends, so drafts imitate you and should be reviewed before anyone treats them as your words. Risk, credentials: The agent acts with your email credentials and permissions, so limit the connector scope and keep approval gates on while you test. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/yc-software-qm/triage-inbox-on-schedule-in-your-voice/ ## How to run tests, open PRs, and watch CI in an existing repo with QM Repo: [yc-software/qm](https://github.com/yc-software/qm) | Difficulty: Advanced | Time: 4 hours | Basis: From the README Have the agent work inside a repository to run tests, open pull requests, and check CI results. **Who it is for:** An engineer on a small team who wants routine repository chores handled while they focus on other work. **The problem:** Running tests, opening PRs, and checking CI after each push is repetitive. Engineers context-switch to do it. **The outcome:** The agent runs the test suite in the repository's sandbox, opens a pull request with the change, and reports the CI status. ### Steps 1. Deploy a QM instance and sign in to the web app. 2. Connect the target repository to the scope's sandbox and grant the agent access to it, since the README says each scope has its own isolated sandbox but does not list the exact connection steps. (inferred, check the README) 3. Ask the agent in the web app or Slack to run the test suite and report failures. 4. Ask the agent to open a pull request with the change and monitor the CI run until it finishes. ### Prompt to paste into your agent ```text Using the repository yc-software/qm, deploy a QM instance and connect the repository to its sandbox. Have the agent run the tests, open a pull request for , and report the CI result. Verify that the pull request exists and that the CI run has a status before reporting success. ``` ### Caveat The README lists this as a supported workflow but does not describe how repository access is granted, so verify the sandbox setup in the deployment docs before relying on it. Risk, credentials: Giving the agent repository access means it can push branches and open pull requests with your permissions, so keep the approval gates on until you trust the workflow. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/yc-software-qm/run-tests-and-open-prs-in-existing-repo/