Kipt

Matt Pocock skills setup: what /setup-matt-pocock-skills does

Matt Pocock skills setup explained: what /setup-matt-pocock-skills asks, which docs/agents files it writes, and how to install and run it once per repo.

By the Kipt team · Updated 24 October 2026

Source: mattpocock/skills · Maintainer: community: Matt Pocock · License: MIT · Checked: 24 October 2026

What is the setup-matt-pocock-skills skill?

Matt Pocock skills setup is a one-time step per repo: you type /setup-matt-pocock-skills, answer two or three questions, and the skill writes a small set of markdown files that tell the other engineering skills where your issues live, what your triage labels are called and where your domain docs sit. It is step 2 of the install instructions in the mattpocock/skills README.

The skill's description reads: "Configure this repo for the engineering skills: set up its issue tracker, triage label vocabulary, and domain doc layout. Run once before first use of the other engineering skills."

It is a real, current skill in the repo, and one of its most installed. As of 24 October 2026, its skills.sh page listed 1.1M installs (snapshot, 24 October 2026).

What does it do?

The SKILL.md calls itself "a prompt-driven skill, not a deterministic script." It runs in five steps:

  1. Explore. It reads git remote -v, any existing CLAUDE.md or AGENTS.md, GLOSSARY.md, docs/adr/, docs/agents/ and .scratch/. It also checks whether the triage skill is installed and looks for monorepo signals such as pnpm-workspace.yaml or a workspaces field in package.json.
  2. Ask, one section at a time, leading each with a recommended answer you can accept in a word.
  3. Show drafts of every file and let you edit them.
  4. Write the files.
  5. Report which skills will now read them.

The three sections:

Section Default it proposes When it asks
Issue tracker GitHub if your remote points at GitHub, GitLab if it points at GitLab Always
Triage labels The five canonical labels: needs-triage, needs-info, ready-for-agent, ready-for-human, wontfix Only if triage is installed
Domain docs Single-context: one GLOSSARY.md plus docs/adr/ at the root Only if it finds monorepo signals

The tracker options are GitHub (uses the gh CLI), GitLab (uses the glab CLI), local markdown under .scratch/<feature>/, or "Other", where you describe a Jira, Linear or similar workflow in one paragraph.

What it writes:

  • docs/agents/issue-tracker.md
  • docs/agents/domain.md
  • docs/agents/triage-labels.md, only when triage is installed
  • An ## Agent skills block in CLAUDE.md if it exists, else AGENTS.md. If neither exists, it asks you which to create.

When the triage labels section ran on GitHub or GitLab, the current SKILL.md also tells it to create any configured label the tracker lacks, using gh label create or glab label create.

Folder contents: SKILL.md, five seed templates (issue-tracker-github.md, issue-tracker-gitlab.md, issue-tracker-local.md, triage-labels.md, domain.md), and agents/openai.yaml for Codex.

Triggers. None automatic. Its frontmatter sets disable-model-invocation: true, and the Codex metadata sets allow_implicit_invocation: false, so it only runs when you type it.

How do you install it?

The README gives a plugin route and an editable-files route, and says to pick one per agent so you don't get every skill twice.

Claude Code plugin (whole collection, updates itself). From Anthropic's official plugin marketplace:

claude plugin install mattpocock-skills@claude-plugins-official

Plugin skills are namespaced as /plugin-name:skill-name, so it appears as /mattpocock-skills:setup-matt-pocock-skills.

Codex plugin. From the README:

codex plugin marketplace add mattpocock/skills
codex plugin add mattpocock-skills@mattpocock

Just this skill, as editable files. The command on skills.sh is:

npx skills add https://github.com/mattpocock/skills --skill setup-matt-pocock-skills

On its own it does little, so add the skills that read its output with extra --skill flags, which the skills CLI supports. The README says that if you run the interactive installer, include setup-matt-pocock-skills when it asks which skills to take. For Claude Code, folders land in .claude/skills/<name>/ in your project, or ~/.claude/skills/<name>/ with -g, the standard Claude Code skill locations. Codex and Cursor read the same SKILL.md format; see Codex skills and Cursor skills.

How do you use it?

Open the repo you want to configure, start your agent and type the command. Most runs need two confirmations.

Example prompts (suggestions, not quotes from the repo):

  • /setup-matt-pocock-skills
  • /setup-matt-pocock-skills We track issues in Linear through the linear CLI, not GitHub.
  • /setup-matt-pocock-skills This is a solo project with no remote. Use local markdown.
  • /setup-matt-pocock-skills Our tracker already uses bug:triage instead of needs-triage.

Requirements. A git repo. For GitHub, the gh CLI; for GitLab, the glab CLI. Local markdown needs nothing, not even a remote.

Check it worked. The setup docs list the signs: docs/agents/issue-tracker.md and docs/agents/domain.md exist, an ## Agent skills section appears in your instruction file, and /to-tickets publishes without asking where issues go. No SKILL.md should have changed.

Which skills read the setup files?

The repo's own ADR on setup pointers splits the engineering skills in two:

Dependency Skills named in the ADR Without setup
Hard to-tickets, to-spec, triage "output is wrong, not just fuzzy"
Soft diagnose, tdd, improve-codebase-architecture Still works; output is less sharp

The domain docs file also tells skills to proceed silently if GLOSSARY.md or ADRs don't exist yet. The domain modeling skill creates them later, when a term or decision is resolved, so an empty repo after setup is normal.

Tips and limits

From the repo's docs, not from testing:

  • It picks the file, not the agent. It edits CLAUDE.md if one exists, even if you are on Codex. Move the block to AGENTS.md by hand, or keep AGENTS.md canonical and point CLAUDE.md at it. See CLAUDE.md and AGENTS.md.
  • No global mode. Every repo carries its own docs/agents/. A user-level option is an open request.
  • Re-run after big updates. The seed templates change between versions. The repo's pending changelog entry for the label and GitLab fixes says to re-run /setup-matt-pocock-skills to refresh docs/agents/.
  • Labels: check your tracker. The current SKILL.md creates missing triage labels, a fix merged after the v1.3.1 release. Older installs and the setup docs page describe a version that does not, so if gh says a label is missing, create it once by hand.
  • It is not a preferences panel. It configures tracker, labels and doc layout only. Personal preferences, such as grilling one question at a time, go in your CLAUDE.md.
  • External PRs are off by default. The GitHub and GitLab templates carry a "PRs as a request surface" flag set to off; flip it in docs/agents/issue-tracker.md if you want triage to cover outside PRs.

The folder holds no scripts. The commands it runs are read-only checks such as git remote -v plus the label commands above, and every file it writes is markdown you can read and edit. It is MIT licensed. The skills.sh audits showed Pass from Gen Agent Trust Hub, Socket and Snyk as of 24 October 2026.

Save it in your Kipt library

In Kipt, open Skills and paste the skill's GitHub folder link into the import box. Kipt reads the SKILL.md and the five seed templates beside it; since the repo holds many skills, it shows a list to pick from, so you can take setup together with triage and the other skills that read its files. You can also drop the folder or a .zip. Your library lives in your own private GitHub repo and installs into Claude Code or Codex with one command:

npx skills add https://github.com/<you>/<repo>.git --all -g

Kipt is not affiliated with Matt Pocock or mattpocock/skills.

Similar skills

Questions

What does /setup-matt-pocock-skills do?

It configures one repo for the engineering skills in mattpocock/skills. It asks where issues live, what your triage labels are called and how your domain docs are laid out, then writes the answers to markdown files under docs/agents/ and adds an Agent skills section to CLAUDE.md or AGENTS.md.

Do I need to run setup-matt-pocock-skills in every repo?

Yes. The README says to run it once per repo, and the docs say there is no user-level or global mode. Each repo gets its own docs/agents/ folder, which you commit like any other markdown.

Do Matt Pocock's skills only work with GitHub?

No. Setup ships templates for GitHub (using the gh CLI), GitLab (using the glab CLI) and local markdown files under .scratch/. For Jira, Linear or anything else, you describe your workflow in a paragraph and setup records it as prose that the other skills follow.

Which skills need setup-matt-pocock-skills?

The repo's ADR calls to-tickets, to-spec and triage hard dependencies: without the config their output is wrong. Skills such as tdd and improve-codebase-architecture only use it to sharpen output and still work without it.

It wrote to CLAUDE.md but I use Codex. What now?

The skill edits CLAUDE.md if it exists, else AGENTS.md, without checking which agent is running. The docs list this as an open gap and suggest moving the Agent skills block to AGENTS.md by hand, or making CLAUDE.md a one-line pointer to AGENTS.md.

More in Skill directory

All guides

Keep your prompts and skills in one place, in your own GitHub. Install every skill with one command.

Get started

Kipt is an independent product, not affiliated with or endorsed by Anthropic, OpenAI, Cursor or GitHub. Claude, ChatGPT, Codex, Cursor and GitHub are trademarks of their owners.