# AI use cases for Factory (addyosmani/factory)

Repo: [addyosmani/factory](https://github.com/addyosmani/factory) | Use cases: 4 | Page: https://dedupedai.com/use-cases/addyosmani-factory/

Factory is a shell-based reference that installs a repeatable, GitHub-backed software delivery loop into an existing project for Claude Code, with a thin Codex adapter.

This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL.

- Install the Factory into an existing GitHub project: https://dedupedai.com/use-cases/addyosmani-factory/install-factory-into-existing-repo/
- Turn small labeled issues into reviewed draft pull requests: https://dedupedai.com/use-cases/addyosmani-factory/turn-labeled-issues-into-draft-prs/
- Sort an issue backlog into implement, spec, or wait: https://dedupedai.com/use-cases/addyosmani-factory/triage-issue-backlog-into-queue-labels/
- Get a weekly health sweep that files issues for findings: https://dedupedai.com/use-cases/addyosmani-factory/weekly-health-sweep-files-new-issues/

## How to install the Factory into an existing GitHub project with Factory

Repo: [addyosmani/factory](https://github.com/addyosmani/factory) | Difficulty: Intermediate | Time: 1 hour | Basis: From the README

Add the committed Claude and Codex Factory files to a repo and check the setup with doctor.

**Who it is for:** A developer who wants to automate issue work in an existing GitHub project and needs the Factory files installed first.

**The problem:** Setting up agent-driven delivery by hand means guessing which policy, skill, label, and gate files a repo needs, and missing one can leave runs silently unsafe.

**The outcome:** The target repo contains the Factory policy, skills, and gates, and the doctor script reports placeholders, missing labels, and remote gaps to fix before any run writes to the repo.

### Steps

1. Clone the addyosmani/factory repository to your machine.
2. Run `install.sh` against the target repository you want to automate.
3. Open the target repository in your coding agent and fill in `docs/factory/CHARTER.md`.
4. Run `.factory/scripts/doctor.sh` and fix every placeholder, missing label, or missing remote it reports.
5. Open `/factory` to confirm the control room shows the expected state before enabling writes.

### Prompt to paste into your agent

```text
Using the addyosmani/factory repository, install the factory into <target-repo-path> by running install.sh against it. Then run .factory/scripts/doctor.sh inside the target and report every finding. Expected output: a list of remaining setup gaps, or confirmation that none remain. Verify that docs/factory/CHARTER.md exists in the target.
```

### Caveat

The README says the Factory is an installer for another project and that the Claude unattended schedules and cloud routines are set up separately, so a successful install does not mean scheduled runs are active.

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/addyosmani-factory/install-factory-into-existing-repo/

## How to turn small labeled issues into reviewed draft pull requests with Factory

Repo: [addyosmani/factory](https://github.com/addyosmani/factory) | Difficulty: Advanced | Time: 2 hours | Basis: From the README

Let scheduled agents implement charter-allowed issues on claimed branches and open draft PRs for review.

**Who it is for:** A maintainer with a backlog of small, well-described bugs who wants them worked overnight while a person keeps the merge decision.

**The problem:** Small fixes sit in the backlog because steering an agent session by hand for each one costs more time than the fix itself.

**The outcome:** Each permitted issue gets a failing test, a scoped change, passing gates, an independent verifier verdict, and a draft PR that a human reviews in the control room before merging.

### Steps

1. Write the scope, risk tier, protected paths, and review-queue limit in `docs/factory/CHARTER.md`.
2. Label a small, clear issue `factory:ready-to-implement`.
3. Let the scheduled implementation routine claim the issue on its deterministic branch and open a draft PR.
4. Wait for the PR verification routine to leave its verdict on the pull request.
5. Open `/factory`, read the evidence in the review queue, then request changes, close, or merge the PR yourself.

### Prompt to paste into your agent

```text
In the addyosmani/factory setup of <target-repo>, pick one issue labeled factory:ready-to-implement and run the implementation stage on it. Expected output: a draft pull request with its gate results and verifier verdict. Verify that no merge was performed and that the PR is still open as a draft.
```

### Caveat

The README says Claude routines do not trigger on newly created issues, so the default hourly triage poll means an issue may wait until the next run, and the optional GitHub Action is needed for immediate triage.

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/addyosmani-factory/turn-labeled-issues-into-draft-prs/

## How to sort an issue backlog into implement, spec, or wait with Factory

Repo: [addyosmani/factory](https://github.com/addyosmani/factory) | Difficulty: Intermediate | Time: 1 hour | Basis: From the README

Have a triage run classify each untriaged issue and write a handoff comment with its gate level.

**Who it is for:** A project lead with a growing pile of untriaged GitHub issues who wants each one routed before anyone starts coding.

**The problem:** Reading every issue to decide whether it is small enough to automate, needs a spec, or lacks a fact takes time and is easy to do inconsistently.

**The outcome:** Each issue carries one queue label and a structured handoff comment with expected files, completion condition, gate level, and confidence, and any issue needing product intent is parked for a human spec.

### Steps

1. Confirm the charter in `docs/factory/CHARTER.md` states what work the Factory may attempt.
2. Run the interactive triage stage in your coding agent against the target repository.
3. Review the labels and the queue snapshot the triage stage produces.
4. Move any issue marked `factory:ready-to-spec` into the human-guided spec workflow with its four approval gates.
5. Open `/factory` to check what now needs you.

### Prompt to paste into your agent

```text
Using the addyosmani/factory skills in <target-repo>, run the triage stage on the open untriaged GitHub issues. Expected output: a list of issues with the factory:* label applied to each and a one-line reason. Verify that no code was changed and that every issue has exactly one queue label.
```

### Caveat

The README describes interactive triage for both Claude Code and Codex, but the unattended hourly triage schedule is supplied only for Claude cloud routines.

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/addyosmani-factory/triage-issue-backlog-into-queue-labels/

## How to get a weekly health sweep that files issues for findings with Factory

Repo: [addyosmani/factory](https://github.com/addyosmani/factory) | Difficulty: Advanced | Time: 2 hours | Basis: From the README

Run a monitor that checks the default branch and Factory health, then files findings as GitHub issues.

**Who it is for:** A maintainer who wants a recurring check on the default branch that turns problems into tracked work without manual searching.

**The problem:** Regressions and Factory setup drift can go unnoticed between releases because nobody has time to scan the branch and the Factory configuration each week.

**The outcome:** Findings from the weekly sweep appear as new GitHub issues that the triage routine can pick up, so the loop closes without a person hunting for problems.

### Steps

1. Install the Factory and confirm `.factory/scripts/doctor.sh` reports no open gaps.
2. Configure the weekly monitor routine supplied by the repository for Claude cloud routines.
3. Let the sweep run and file its findings as GitHub issues.
4. Check the new issues in `/factory` and confirm the triage routine labels them.
5. Adjust constraints with `/factory-tune` when the monthly review suggests changes.

### Prompt to paste into your agent

```text
For the factory installed in <target-repo>, run the monitor stage once and report the issues it would file. Expected output: a list of proposed issue titles and the default-branch or setup finding behind each. Verify that the stage lists findings before filing any issue.
```

### Caveat

The README says the Factory has no background process and that each routine starts fresh and stops, so the weekly sweep only runs when its Claude routine is configured and scheduled.

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/addyosmani-factory/weekly-health-sweep-files-new-issues/