mattpocock-skills
claudeMatt Pocock's agent skills for real engineering (grilling, spec/ticket flows, TDD, code review, domain modelling and more).
39 skills
Distills lessons from your real projects and injects the relevant ones back at the right moment. Ships 12 specialist agents (incl. a frontend builder and an authorized Red Team pentester), a recall skill, and the /brain review flow. Requires the `raph` CLI: npm install -g maheshaggarwal21/raphael.
Plugin raphael-brain from marketplace raphael in maheshaggarwal21/raphael. Commands read from the repo's manifest — 1 month ago.
This plugin runs code on your machine — its hooks fire automatically on events like tool use. The exact commands are listed below; read them before installing.
brain-recallPull relevant lessons from the developer's Raphael brain when stuck, debugging a repeated failure, or before risky changes (deploys, migrations, auth, payments, webhooks).
/raphael-brain:brain-evalProve the brain helps — run the eval harness (canaries + ON/OFF lift)
/raphael-brain:brain-learnMine your recent sessions and distill them into candidate lessons
/raphael-brain:brain-reviewReview candidate lessons — approve, reject, or edit the queue
/raphael-brain:brainRaphael brain — status, setup, and what to do next (start here)
READMEraphael-architectthe senior systems architect who designs a premium, scalable architecture from the spec. Use this agent proactively when a spec or feature is agreed and needs a technical design, data model, or system structure before anyone writes implementation code. (Raphael agent)
raphael-critiquethe adversarial pass over any other agent's output before you see it. Use this agent proactively when another agent's output, a plan, or a confident claim should be stress-tested for unsupported reasoning before the user relies on it. (Raphael agent)
raphael-debuggerthe production-grade root-cause finder. Use this agent proactively when something is broken, throwing, failing a test, or behaving wrong and the root cause is not obvious — use PROACTIVELY the moment an error or unexpected behaviour appears. (Raphael agent)
raphael-deployerpre-ship checks: migrations, env vars, rollback. Use this agent proactively when the user is about to ship, deploy, or release and needs a pre-flight checklist — it produces the plan and STOPS, never deploys. (Raphael agent)
raphael-designreviews UI/UX and visual consistency. Use this agent proactively when a UI needs a taste and accessibility review, looks generic or inconsistent, or the user asks whether a design is any good. (Raphael agent)
raphael-developerwrites code with the relevant past lessons already in context. Use this agent proactively when it is time to implement agreed backend or general (non-UI) code changes against a plan or a concrete task. (Raphael agent)
raphael-frontendbuilds distinctive, non-generic UI — the frontend where AI lags most. Use this agent proactively when it is time to BUILD or reshape UI — a landing page, component, screen, or design system — and you want output that does not read as templated "AI slop". (Raphael agent)
raphael-managerthe router that turns your request into the right specialists and merges their results. Use this agent proactively when a request spans several steps, or you are not sure which specialist it needs — this agent routes to the right ones and merges their answers. (Raphael agent)
raphael-plannerthe idea improver / finaliser who turns a vague idea into a sharp, buildable spec. Use this agent proactively when the user has a fuzzy idea, feature request, or "let's build X" with no clear spec yet — run this BEFORE any design or code. (Raphael agent)
raphael-redteamthe attacker's-eye penetration tester that tries to actually break a system you own, then reports what's exploitable. Use this agent proactively when the user wants an authorized attacker's-eye penetration test of THEIR OWN app or a test/staging environment — actively probing for exploitable auth by
raphael-reviewerreviews a diff like a senior engineer who just joined the codebase. Use this agent proactively when a diff, branch, commit, or uncommitted change is ready and should be reviewed before merging or shipping — use PROACTIVELY before any merge. (Raphael agent)
raphael-securityaudits for secrets, injection, and auth mistakes. Use this agent proactively when code touching auth, payments, user data, secrets, file uploads, or input handling is being written or shipped — a DEFENSIVE static audit of the code. (Raphael agent)
raph inject --event session-start
raph inject --event user-prompt
raph inject --event pre-tool
raph subagent-stop
raph pulse --async
Matt Pocock's agent skills for real engineering (grilling, spec/ticket flows, TDD, code review, domain modelling and more).
39 skills
Persistent memory for Claude Code: the next session starts where the last one left off.
29 skills · 1 MCP server · hooks
AI-powered codebase understanding — analyze, visualize, and explain any project
9 skills · 10 agents · hooks
Open-source AI job search that runs locally in your AI coding CLI: scan job portals, evaluate listings into a structured A-H report with a 1-5 score, tailor your CV, track applications
5 skills · hooks
Cross-session memory and token savings for coding agents.
6 skills · 1 agent · 1 MCP server · hooks
Rewrite AI-sounding text so it reads like the writer without changing what it says.
1 skill
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.