mattpocock-skills
claudeMatt Pocock's agent skills for real engineering (grilling, spec/ticket flows, TDD, code review, domain modelling and more).
39 skills
Pre-PR review and structural-smell habits built on the Foundry six-verb gate: one rule set enforced in-loop, pre-push and in CI.
Plugin foundry from marketplace cmaintz in CMaintz/cmaintz-skills. Commands read from the repo's manifest — 6 days ago.
This plugin runs code on your machine — its hooks fire automatically on events like tool use and it starts local MCP servers. The exact commands are listed below; read them before installing.
codemapMap a repo into a committed, cited, SHA-stamped picture of the codebase under docs/codemap/ (an INDEX plus one file per domain slice and three lenses), so every session starts from the same verified map; `refresh` re-audits only the slices whose cited files changed. It describes, it never judges, an
featureDrive a feature ticket to a reviewable PR through the Foundry gate - claim it, implement in an isolated worktree, loop against `mise run gate` until green, verify against the acceptance criteria, then hand to `ship`. Invoke with a ticket id/ref to work that ticket; invoke with no argument to claim t
foundry-secretTriage a secret-scanner finding (gitleaks) and act on it - allowlist a confirmed false positive, or guide rotate-then-purge for a real leaked secret. Use when a secret scan flags something, or the user asks to handle a gitleaks/secret finding.
loop-reportMake the agent loop legible - read Foundry's verb telemetry to show where the gate spends time, whether the loop is converging or thrashing (a verb that passed then failed again), and how many gate attempts a change took; optionally fold in CI outcomes via `gh run`. Use when the user asks why the lo
pr-ci-watchLoop a PR's CI result back to a working agent WITHOUT blocking it. Run the PR's checks watch as a hidden Claude Code background task while the main agent keeps coding, then read the verdict and, on failure, fix, push and re-watch until green. Use right after you open or push to a PR and want CI watc
repo-alignRun an incremental campaign to bring a repo up to the Foundry standard - consistent formatting and no unjustified structural smells - one small reviewable PR at a time. Use when the user wants to pay down code-quality debt across a codebase gradually, or asks to "align the repo", "grind down the sme
reviewReview a diff, branch, or PR across correctness, standards, and spec lenses (plus a red-team lens at high effort), then try to refute every finding before reporting it. Scaled to the size of the change and tunable by effort (low|medium|high). One reviewer to invoke instead of juggling three. Use whe
shipTake a finished change from working tree to pull request through the Foundry gate - auto-fix, structural smells, a green gate, a fresh-context review and verify against the linked issue, a conventional commit, then the PR. Invoke this to land ANY finished change: run it BEFORE you `git commit`/`git
specTurn a rough idea into a concrete, testable spec through a short back-and-forth, then file it as an `agent-feature` GitHub Issue (`agent:ready`) that `feature` can claim. The front of the Foundry loop - idea in, groomed ticket out. It does not implement. Triggers: brainstorm/scope/spec out a feature
verifyCheck a change against its ticket - every acceptance criterion and every Definition of Done item, including DoD items outside the diff - in a fresh-context, read-only subagent that returns pass or fail with evidence per item. Advisory like `review`; it blocks nothing. Use when the user says "verify"
foundryruns: node ${CLAUDE_PLUGIN_ROOT}/scripts/explain-mcp/server.mjs
bash "${CLAUDE_PLUGIN_ROOT}"/hooks/guard-generated-files.sh
bash "${CLAUDE_PLUGIN_ROOT}"/hooks/guard-pr-create.sh
bash "${CLAUDE_PLUGIN_ROOT}"/hooks/guard-governance-labels.sh
bash "${CLAUDE_PLUGIN_ROOT}"/hooks/auto-format.sh
bash "${CLAUDE_PLUGIN_ROOT}"/hooks/habit-hooks-guard.sh
bash "${CLAUDE_PLUGIN_ROOT}"/hooks/typecheck-stop.sh
bash "${CLAUDE_PLUGIN_ROOT}"/hooks/format-java-stop.sh
bash "${CLAUDE_PLUGIN_ROOT}"/hooks/jev-toolcall-triage.sh
bash "${CLAUDE_PLUGIN_ROOT}"/hooks/leash.sh hook
bash "${CLAUDE_PLUGIN_ROOT}"/hooks/leash.sh snapshot
bash "${CLAUDE_PLUGIN_ROOT}"/hooks/leash.sh session
sh "${CLAUDE_PLUGIN_ROOT}"/hooks/mcp-deps.sh
Matt Pocock's agent skills for real engineering (grilling, spec/ticket flows, TDD, code review, domain modelling and more).
39 skills
CMEM Pro for the Claude app - hooks stream tool use to cmem.ai and inject observations into new sessions and agents
3 skills · 1 MCP server · hooks
AI memory system — mine projects and conversations into a searchable palace. 47 MCP tools, auto-save hooks, guided setup.
3 skills · 6 commands · 1 MCP server · hooks
Use Alibaba Open Code Review from local Codex.
2 skills
Universal memory runtime for Claude Code with cross-session memory, entity-aware recall, lifecycle hooks, skills, commands, and memory-curator agents
7 skills · 7 commands · 2 agents · hooks
Long-term semantic memory for Claude Code, powered by OpenViking
5 skills · 1 command · 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.