Guide · Agent skill
How to use the Setup Matt Pocock Skills agent skill
Updated · written with AI from the project's own documentation and checked against our catalogue data
setup-matt-pocock-skills is an agent skill from mattpocock/skills that configures a repository for Matt Pocock's engineering skills. It records where the repo's issues live, which triage label names to use, and how domain docs are laid out. This guide covers what the skill does, how to install it, how to run it with a plain request, and what it needs and touches, so you can run it once before using the other engineering skills.
Setup Matt Pocock Skillsmattpocock/skills · MIT Download the skill Quick start
- 1Install the skill. Download the ZIP from this page and unzip it into ~/.claude/skills/, or into the project's .claude/skills/ folder. In the Claude apps, you can instead upload it under Settings → Capabilities → Skills.
- 2Open the repository. Start the agent from inside the repo you want to configure, so it can read the git remote and existing files.
- 3Ask for the setup by name. Request the setup-matt-pocock-skills skill in plain language. The skill does not start automatically.
- 4Choose an issue tracker. Accept the recommended tracker, or pick GitHub, GitLab, local markdown or another tool. The skill confirms before it writes.
- 5Decide on triage labels. If the triage skill is installed, keep the default labels or give the names your tracker already uses.
- 6Review the drafts. Read the Agent skills block and the docs/agents/ files the skill proposes, and edit them before they are written.
- 7Confirm the tracker works. Run gh repo view or glab repo view from the repo folder to check that the project is reachable.
What does the setup skill make the agent do?
The skill inspects the current repository first. It checks the git remote, whether AGENTS.md or CLAUDE.md exist, and whether glossary and ADR folders are already present. It then walks you through choosing where issues are tracked: GitHub, GitLab, local markdown, or another tool you describe.
After you confirm, it writes configuration notes into CLAUDE.md or AGENTS.md and into docs/agents/. Those notes tell the other engineering skills how to work with your repo. The skill asks for confirmation before it writes anything.
- Issue tracker: where issues live (GitHub by default; local markdown is also supported)
- Triage labels: the strings used for the five canonical triage roles, asked only if the triage skill is installed
- Domain docs: where GLOSSARY.md and ADRs live and the rules for reading them
How do you install the setup-matt-pocock-skills skill?
You can add the skill in one of two ways. The first is to download the ZIP from this page and unzip it into a skills folder. Use ~/.claude/skills/ to make the skill available across your projects, or a project's .claude/skills/ folder to limit it to one repo.
The second is to upload the ZIP in the Claude apps through Settings → Capabilities → Skills.
- Personal install: unzip into ~/.claude/skills/
- Project install: unzip into the project's .claude/skills/ folder
- Claude apps: Settings → Capabilities → Skills, then upload the ZIP
How do you trigger the skill with a natural request?
This skill is set to manual invocation (disable-model-invocation is true in its SKILL.md), so the agent will not start it on its own just because your task looks related. Ask for it by name in plain language, from inside the repository you want to configure.
Run it once before you first use the other engineering skills, such as to-tickets, triage, or to-spec, which read the files it writes.
- Run the setup-matt-pocock-skills skill for this repo.
- Set up this repo for the engineering skills: issue tracker, triage labels and domain docs.
- Configure this repository's issue tracker before we use the triage skill.
What happens during a setup run?
The run follows a fixed order. The agent explores the repo, presents what it found, and asks one question per section. Each section opens with a recommended answer, so you can accept it with a single word. A section is skipped when exploration already settled it.
For the issue tracker, the skill proposes GitHub if a git remote points at GitHub, and GitLab if it points at gitlab.com or a self-hosted GitLab. Otherwise it offers GitHub, GitLab, local markdown under .scratch/, or another tool described in one paragraph.
For domain docs, it defaults to a single-context layout with one GLOSSARY.md and docs/adr/ at the repo root. It offers a multi-context layout only when monorepo signals are found. Before writing, it shows you a draft of each file so you can edit it.
- Step 1, Explore: read git remote, AGENTS.md, CLAUDE.md, glossary files, ADR folders, docs/agents/, and .scratch/
- Step 2, Ask: confirm the issue tracker, then triage labels if triage is installed, then domain layout
- Step 3, Confirm: review and edit drafts before anything is written
- Step 4, Write: update the chosen agent file and the docs/agents/ files
What files does the setup create or change?
The skill edits whichever of CLAUDE.md or AGENTS.md already exists. If neither exists, it asks which one to create rather than choosing for you. If an Agent skills section is already there, it updates that section in place instead of adding a duplicate.
The docs it writes depend on your answers. The skill uses its seed templates for GitHub, GitLab, and local-markdown trackers, and it writes issue-tracker.md from scratch for other trackers. It writes triage-labels.md only when triage is installed.
- CLAUDE.md or AGENTS.md: an Agent skills section with issue tracker, triage labels and domain docs summaries
- docs/agents/issue-tracker.md: where issues are tracked and how to work with them
- docs/agents/triage-labels.md: label mapping (only when triage is installed)
- docs/agents/domain.md: how engineering skills should read GLOSSARY.md and ADRs
What does the skill need, and what are the safety notes?
The skill needs git to read the remote configuration. For a GitHub tracker it needs the gh CLI, and for a GitLab tracker it needs the glab CLI, both authenticated to the host. It also needs a repository remote for GitHub or GitLab, either one that already exists or one you supply. It runs scripts and uses the network.
Some steps change a live project. With your confirmation, the agent can run gh or glab commands and create the configured labels in your tracker. It can also run git init and git remote add origin <url> if the repo has no remote for the host you chose. If you do not want the agent to run these commands or create labels, do not use this skill, or review each step before approving it.
- git: required to read the remote configuration
- gh CLI: for GitHub trackers, authenticated to the host
- glab CLI: for GitLab trackers, authenticated to the host
- Network access: used by the run
- Label creation: gh label create or glab label create for labels the tracker lacks
- Repository setup: git init and git remote add origin <url> if needed
gh repo viewHow do you check that the setup worked?
For GitHub or GitLab trackers, the setup is done when the repo command prints your project from inside the repository folder. If it does not, the skill stops and reports what is missing. Run the matching command yourself to confirm.
After setup, you can edit the files in docs/agents/ directly. Re-run the skill only if you want to switch issue trackers or start over.
- GitHub: run gh repo view from the repo folder
- GitLab: run glab repo view from the repo folder
glab repo viewWhat people use it for
- Setting up a new repo to use the to-tickets, triage, or to-spec engineering skills with a consistent issue tracker
- Recording a team's triage label convention, such as mapping needs-triage to an existing bug:triage label
- Configuring a solo project with local markdown issues under .scratch/ when there is no remote
- Adding a documented layout for GLOSSARY.md and ADRs so later skills read domain terms the same way
Frequently asked questions
How do I install the setup-matt-pocock-skills skill?
Download the ZIP from this page and unzip it into ~/.claude/skills/ for a personal install, or into a project's .claude/skills/ folder for one repo. In the Claude apps, you can upload the ZIP under Settings → Capabilities → Skills.
Is setup-matt-pocock-skills free to use?
The skill is published under the MIT licence, as listed on this page. The source does not mention pricing, so check the licence and the repository for any other terms.
What do I need before running the skill?
You need git to read the remote configuration. For GitHub you need the gh CLI, and for GitLab the glab CLI, both authenticated to the host. You also need a repository remote for that host, either existing or supplied during setup.
Why doesn't the agent run the setup skill automatically?
The skill is set to manual invocation, so the agent will not start it from a general request. Ask for it by name, and run it once before you use the other engineering skills.
Can I skip the GitHub or GitLab setup and use local markdown instead?
Yes. Local markdown stores issues as files under .scratch/<feature>/ in the repo. It suits solo projects and repos without a remote, and the skill offers it as one of the tracker choices.
What if the skill can't connect to my GitHub or GitLab project?
The skill checks that gh repo view or glab repo view prints the project from inside the repo. If it can't, it stops and tells you what is missing, such as a missing remote or authentication. Fix that and run the check again.
New agent skills, repos and AI tools — every week
The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.
Our verdict
Use this skill if you are setting up a repo for Matt Pocock's engineering skills and want a shared issue tracker and triage label convention written into your repo's docs. Pick something else if you will never use those engineering skills, since the configuration does nothing on its own, or if you do not want the agent running gh or glab commands or creating labels in your tracker.
Download the skill