mattpocock-skills
claudeMatt Pocock's agent skills for real engineering (grilling, spec/ticket flows, TDD, code review, domain modelling and more).
39 skills
A team of AI agents that builds software end to end, with a shared handoff contract and approval gates.
Plugin software-factory from marketplace software-factory-marketplace in achisolomon/open-software-factory. Commands read from the repo's manifest — 2 days 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.
factory-initSet up the current repository for the Software Factory agent team — create the docs/ scaffold, capture project facts, and (for existing code) map the repo and derive coding standards. Run once per project.
frontend-testingPlan, write, debug, and verify browser tests for web UI changes using the existing Test Automation Engineer. Use for page or component behavior, responsive layouts, keyboard accessibility, and browser failures; skip for backend-only work and ordinary unit tests.
healthRead-only health check of a Software Factory project. Flags stale in-flight dispatches, dead background processes, worktrees with uncommitted or unpushed work, board drift and errors in logs. Use when the user asks for a factory health check, or before a long run or a cleanup.
resumeResume the Software Factory Team Lead after a context compaction or a session restart. Reconciles every in-flight dispatch against the reports on disk before any new work is sent.
reviewRun a full team review of the current changes, a branch, a task or the whole codebase — architecture and clean-code review, security review, test verification, and cost review when infrastructure changed — and consolidate one verdict.
session-setupConfigure the project's Claude Code settings for Software Factory runs. Sets the auto-compact threshold and an allowlist of exact read-only and stack commands, after showing the diff and asking. Claude Code only.
statusShow the Software Factory Team table now - who is running, blocked, done since the last status, and what is next. Read-only. Use when the user asks for the status, the team table, or what the team is doing.
team-leadUse when the user says "team lead" or "team leader", addresses the team lead, or asks the team or factory to build, fix, review or report status. The current session becomes the Software Factory Team Lead and drives the specialist agents, stopping at the approval gates.
devops-engineerDevOps Engineer of the Software Factory. Use it to plan environments and the deployment model during design review, and to build containers, CI/CD pipelines, infrastructure as code, observability (logs, metrics, traces, alerts), runbooks and release automation. Validates with plans and dry runs only
exploratory-qa-engineerExploratory QA Engineer (black-box bug hunter) of the Software Factory. Use it at epic end, or when the user asks, to use the running product the way a user would, put data into odd states on a local stack, and prove each bug with a failing repro test. Not a per-task reviewer. Works on local or disp
finops-engineerFinOps Engineer of the Software Factory. Use it to estimate the cost of an architecture before it is built, set cost guardrails, review tasks and infrastructure-as-code plans that add or change paid resources or third-party APIs, check CI workflow changes and the GitHub Actions budget (minutes, larg
performance-engineerPerformance Engineer (load tester and sizing expert) of the Software Factory. Use it to review the architecture and code for what breaks under load (missing indexes, N+1 queries, lock ordering and deadlocks, unbounded queries and fan-out, pool and queue limits, synchronous work on hot paths), to siz
product-managerProduct Manager of the Software Factory. Use it to turn an idea, feature request or problem into a product vision, a PRD with measurable success metrics and non-functional requirements, and a prioritized backlog of epics and user stories with testable acceptance criteria; and at release time to veri
project-managerProject Manager of the Software Factory. Keeps the GitHub issues and the GitHub Project board in step with the work. Use it to set up the board, to open the issues that are missing (stories, tasks, findings, follow-ups), to move each issue's Status and post its comments as work is dispatched, review
security-engineerSecurity Engineer of the Software Factory. Use it to threat-model the architecture and set security requirements during design review, to review task diffs, new dependencies and infrastructure-as-code for vulnerabilities, to run read-only security scans (SAST, dependency, secrets, container, IaC), a
software-architectSenior software architect and code reviewer for any project. Use it (1) to map an existing repo or system into Mermaid diagrams, (2) to produce or update the system design — application structure, data flow, component interactions and deployment model — and (3) to review code, pull requests or a who
software-developerSoftware Developer of the Software Factory. Use it to implement one task from docs/architecture/tasks.md at a time — application code plus its unit tests, following the architecture, ADR rules and coding standards — and to fix findings from architecture, security, test and cost reviews. Works on a l
team-leadTeam Lead and orchestrator of the Software Factory agent team. Runs as the main session (claude --agent software-factory:team-lead, or via /software-factory:build). Takes a product idea, feature or fix, drives it through discovery, architecture, build, review and release by classifying each request
test-automation-engineerTest Automation Engineer of the Software Factory, and the team's devil's advocate. Use it to write the test strategy from the PRD and architecture, build and maintain the test framework, write integration, contract and end-to-end tests, verify each task against its acceptance criteria, hunt for the
triage-classifierTriage Classifier of the Software Factory. Use it first on every new request the Team Lead receives (except a plain status question) to classify the work - type, size, the lifecycle path, which conditional reviewers are needed, and whether it is clear enough to start - so small work exits early and
ux-ui-designerUX/UI Designer of the Software Factory. Use it to turn PRD stories into user flows, screen-by-screen wireframe descriptions and a design-tokens system during Discovery and design review, to review the Architect's design and each Build-loop task's UI diff for usability and accessibility, and to hold
python3 "${CLAUDE_PLUGIN_ROOT}/hooks/require_agent_model.py"
python3 "${CLAUDE_PLUGIN_ROOT}/hooks/project_bounds.py"
python3 "${CLAUDE_PLUGIN_ROOT}/hooks/gate.py"
python3 "${CLAUDE_PLUGIN_ROOT}/hooks/no_wait.py"
python3 "${CLAUDE_PLUGIN_ROOT}/hooks/check_reply.py"
python3 "${CLAUDE_PLUGIN_ROOT}/hooks/version_check.py"
python3 "${CLAUDE_PLUGIN_ROOT}/hooks/scratch.py" start
python3 "${CLAUDE_PLUGIN_ROOT}/hooks/after_compact.py"
python3 "${CLAUDE_PLUGIN_ROOT}/hooks/scratch.py" end
python3 "${CLAUDE_PLUGIN_ROOT}/hooks/team_lead_trigger.py"
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
Cross-session memory and token savings for coding agents.
6 skills · 1 agent · 1 MCP server · 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
Automates browser interactions for web testing, form filling, screenshots, and data extraction
12 skills
Use Alibaba Open Code Review from local Codex.
2 skills
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.