natterjack
claudecodexSet up Natterjack for your team: a multi-repo Claude Code workspace with a local dashboard (runs, issues, apps, machine setup, docs, links) tailored to your stack, tools and brand.
6 skills
Workflow skills for multi-repo workspaces: implement a ticket end to end with subagents, align on it first, worktrees per ticket, safe pulls, commits and cross-linked PRs across every repo, a local Copilot second-opinion review, and plans posted to your issue tracker.
Plugin multirepo from marketplace lhoezee-skills in lhoezee/skills. Commands read from the repo's manifest — today.
alignReach explicit shared understanding with the engineer on a ticket's problem and solution BEFORE any planning or code. Fetches the ticket (Linear, Jira or GitHub Issues), flags terms and phrasing likely to be misread, then runs a one-question-at-a-time Socratic interview (terminology, problem, a quic
approveThe software factory's Approver. Rate a ticket's change for risk (the team's deterministic risk rules first, then a conservative judgement of the ticket, plan, diff and review results), record the assessment on the Natterjack Factory board, and approve or merge only when the team's policy enforces i
commitCreate Conventional Commits across every repo of a multi-repo workspace (or a single repo), with the ticket from the branch name, context from the issue tracker, and a message proposed from the actual diff. Use whenever the user wants to commit, "commit this", "commit my changes", "make a commit", "
copilot-reviewGet a local second-opinion code review from the GitHub Copilot CLI (a different model, e.g. GPT) on your local changes BEFORE pushing or opening a PR. Use this whenever the user wants a pre-push review, a "second pair of eyes" / "second opinion" on a diff, to "have Copilot/GPT look at my changes", t
factoryRun the software factory from a terminal session. Pull the tickets the team's Factory policy takes in (by tracker state and label), up to its WIP limit, and run each through implement under the policy, one at a time, each in its own worktree. Use when the user wants to "run the factory", "pull work
fetch-ticketFetch an issue-tracker ticket (Linear, Jira or GitHub Issues) with everything that matters for building it (description, acceptance criteria, custom fields, linked/blocking issues, subtasks, parent, and every comment with author and date, including plans and alignment docs other skills posted) and s
implementImplement a ticket (Linear, Jira or GitHub Issues) end to end across a multi-repo workspace (or a single repo) - find or create the ticket's worktree, check the repos are current, settle open questions up front, branch every affected repo, then run a pipeline of subagents in dependency order that pl
prOpen pull requests for every repo a ticket touched (in dependency order, cross-linked, ticket in the title), update the ticket in Linear, Jira or GitHub Issues, and optionally request a Copilot review and watch the PRs, fixing valid review feedback, replying to and resolving threads, and keeping bra
pullBring the workspace's local repos (and the workspace root itself) up to date with origin's default branch, safely - fetch everything, fast-forward only the repos that are clean, on their default branch and behind, and report every other repo with the reason. Never switches branches, stashes or merge
retroThe software factory's retro for a ticket whose PRs merged. Record what review caught, which of the plan's assumptions were wrong, and how the Approver's call held up (reverted? hotfixed?), on the Natterjack Factory board, and propose CLAUDE.md, skill or risk-rule tweaks as a follow-up for a person
review-guideWrite a guided review for a pull request (the order to read it in: intent, the core change, its consequences, the glue, risks, and what to check) and post it as one comment on the PR, updated in place on later runs. The Natterjack dashboard's Reviews page shows the same guide. Use when the user want
spec-checkScore a ticket's spec readiness (out of 100) the same way the software factory's spec gate does, with the team's thresholds and weights from the Natterjack Factory policy, and say what's missing. Works on a draft that isn't in the tracker yet, or on an existing ticket. Use when someone is writing or
submit-planPost an implementation plan from plan mode (~/.claude/plans/) to an issue-tracker ticket (Linear, Jira or GitHub Issues) as a tagged comment, updating that same comment on later submissions instead of adding duplicates. Use whenever the user wants to push, submit, sync, attach or update a ticket wit
tddRed/green/refactor test-driven development discipline for agents writing production code. Use whenever implementing a feature, a business rule, a validation, a calculation or a bug fix in any language: write one failing test, prove it fails for the right reason, write the minimum code to pass, run t
worktreeCreate, list, remove, purge and sync isolated per-ticket worktrees for parallel work across a multi-repo workspace (or a single repo). Use whenever the user wants a separate copy of the repos to work on a ticket without disturbing the main checkout, e.g. "new worktree for ENG-123", "spin up a worksp
Set up Natterjack for your team: a multi-repo Claude Code workspace with a local dashboard (runs, issues, apps, machine setup, docs, links) tailored to your stack, tools and brand.
6 skills
Matt Pocock's agent skills for real engineering (grilling, spec/ticket flows, TDD, code review, domain modelling and more).
39 skills
Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques
15 skills · hooks
Up-to-date documentation lookup. Pull version-specific documentation and code examples directly from source repositories into your coding agent's context.
1 MCP server
Use Alibaba Open Code Review from local Codex.
2 skills
Claude Code skill for creating and presenting Slidev slidedecks
1 skill
Code intelligence powered by a knowledge graph. Provides execution flow tracing, blast radius analysis, and augmented search across your codebase.
12 skills · 1 MCP server · hooks
This page updates itself from your repo. Add the badge to your README, or get it Featured to pin it above the organic results with a followed link to your site.