Agentic Skills & Commands

Resolve PR Review Comments

One Front Door for Agent Architecture, Config, and Setup

Which Dev Loop skill to run

Board

Prune Merged Branches, Stale Worktrees, and Finished Work

Execute one prepared issue

Single Front Door for Deployment and Infra Provisioning

UI/Design Review and Refinement

Environment Variable Management

Prepare an End-to-End Feature

Diagnose and Fix Failing Checks

Execute one prepared issue

Merge ALL Approved Open PRs Into the Trunk

Pull Request Lifecycle

Prepare Work for Agents

Optimize a Prompt

Structured Verification Pass on Your Work

One Front Door for Improving Existing Code

One Front Door for Every Code Review

ICP → revenue-ranked backlog → dated milestones

Run a Security Audit

One Front Door for Authoring and Maintaining Agent Skills

Post Inline Suggested Changes on a PR

One Front Door for Running, Authoring, and Setting Up Tests

Applies WCAG 2.1 AA to web UI and fixes failures

Audits LLM agent apps for wrapper regressions, prompt or memory contamination, tool misuse, and hidden retry loops

Audits agent instruction files (AGENTS.md, CLAUDE.md, hooks, settings) across workspaces, report-only

Router behind /agent

Adds or repairs .agents/ project context in an existing repo

Audits and cuts AI agent token and inference spend via context discipline, prompt caching, model routing, and batching

Designs AI waiting-state UX: thinking indicators, reasoning display, progress steps, and streaming states

Designs RESTful APIs for NestJS

Sketches types, signatures, and module structure before code, then guides implementation

Runs N parallel candidates at the same task, picks a base, and grafts the best parts of the others into it

Builds multi-component claude.ai HTML artifacts with React, Tailwind, and shared UI. Use for complex artifacts needing state management, not simple single-file HTML

Ask which Dev Loop skill or flow fits the current situation

Sweeps frontend accessibility, performance, theming, responsive design and implementation integrity in one scored P0-P3 report

Sets up AWS infrastructure: EC2, VPC, security groups, load balancers, Route53 DNS, and ACM certificates

Finds what a small change could break elsewhere and proves it with real code

Reconciles a project board with current work and delivery evidence, reporting gaps; applies only approved field changes

File a GitHub issue of type Bug from a description

Improves unclear UX copy, error messages, microcopy, labels, and instructions

Reviews a PR or diff for correctness, security, and spec fidelity, including TypeScript, database safety, tests, and flag leaks

Supplies deep-module vocabulary and runs read-only codebase surveys (plans, report, deepen)

Generates raster images (icons, illustrations, textures) via the logged-in Codex CLI image tool

Granular feedback on drafts without rewriting

Writes Conventional Commit messages from git changes, splits unrelated changes, flags breaking changes, and commits after approval

Guides React/Next.js component architecture for monorepos

Builds browser-extension content scripts

Improves context efficiency and diagnoses context degradation

Generates a project-local verification skill that drives the app like a user

Critiques design from a UX view

Advises engineering leadership

Debugs failures end to end: builds a repro loop, ranks hypotheses, instruments, and fixes the root cause with a regression test

Audits the dependency supply chain

Runs deployment workflows for web apps to staging or production

Router behind /deploy

Composes deployment workflows from repo signals

Hunts design-token drift in frontends

Router behind /design

Strips AI slop from code, product, and prose

Scaffolds a VS Code Dev Container with Docker, docker-compose, and optional Claude Code CLI. Use when setting up a devcontainer or containerized dev environment

Writes Dockerfiles, multi-stage builds, and docker-compose setups for NestJS and Next.js apps

Writes or updates README files, guides, runbooks, API references, and setup docs matching the repo's conventions

Builds and sharpens a project's domain model

Deploys backends to EC2 via Docker, GitHub Actions CI/CD, and Tailscale SSH. Use when wiring automated deploys for NestJS, Next.js, or Express to EC2

Finds the env vars a codebase reads, generates .env.example, validates .env, and checks .gitignore covers secrets

Designs error handling, exception filters, error responses, logging, and recovery for React, Next.js, and NestJS. Use when implementing or reviewing error flows

Builds evaluation frameworks for agent systems, including LLM-as-judge design

Executes a prepared GitHub issue without inventing decisions, escalates plan gaps, and tracks review, CI, merge, and deploy evidence

Scaffolds a runnable Expo React Native app with screens, Expo Router navigation, and optional Better Auth

Turns a rough request into one execution-ready issue via the shared requirements, planning, and issue-publishing engines

Designs an auditable playbook when no narrower one fits

Finishes an implementation branch through authorized PR publication and verified delivery gates, or preserves/discards it when explicitly requested

Resolves git merge conflicts correctness-first, then proves the tree still builds

Builds distinctive, production-grade frontend interfaces that avoid generic AI aesthetics

Reviews PR structure, security depth, and devex/flag hygiene with parallel agents after code-review passes correctness

Audits candidate code and intent against current trunk, then plans or removes only proven-safe or merged branches and worktrees

Guards daily git work: blocks staged secrets, gates destructive git commands, installs pre-commit hooks, drives leak response

Author, review, and harden GitHub Actions workflows using current official documentation, secure trigger patterns, least-privilege permissions, current action versions, and CI/CD validation

Diagnoses failing or stuck GitHub Actions checks on a PR, finds root cause, and applies targeted fixes, optionally looping until green

Triages a GitHub work inbox of assigned issues, review requests, mentions, and authored PRs with failing checks

Creates, updates, and publishes GitHub pull requests with clean titles, durable bodies, branch hygiene, and safe push gates

Reviews GitHub PRs and posts inline comments with applyable suggestion blocks through gh

Designs and reviews GraphQL schemas, resolvers, mutations, pagination, and data loading

Grills the user relentlessly about a plan, decision, or idea until every branch of the design tree is resolved

Runs an independent second-opinion review of a diff through the Grok CLI and verifies every finding against the code

Writes a handoff document so a fresh agent session can continue the current work

Explains how a subsystem works

Applies opinionated styling to barebones HTML, turning tables, lists, status text, buttons, and layouts into a cohesive design

Sets up or verifies Husky pre-commit hooks that enforce test coverage above 80% (configurable) for Node/TypeScript projects using Vitest (or Mocha with nyc/c8)

Documents a product's Ideal Customer Profile as .agents/memory/icp.md, with segments ranked by revenue potential

Builds resilient data ingestion pipelines from paginated APIs using two watermarks so progress is never lost

Runs several independent adversarial reviewers over a diff and synthesizes one verdict without auto-applying fixes

Repo-grounded discovery interview that produces a handoff brief for PRD writing, feature intake, or planning

Scaffolds a production-ready static landing page with working email capture form, analytics, and responsive design

Reworks the structure under a UI

Sets up Biome (default) or ESLint + Prettier, Vitest, and pre-commit hooks for JavaScript/TypeScript projects with Bun

Designs prompts, schemas, validation, and retry logic for reliable machine-readable model output

Keeps a project's verification skill and feature map accurate using parallel source readers and one live session, ending in at most one PR of proven corrections

Creates MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools

Designs memory architectures for agents that persist state across sessions and keep entities consistent

Review and land open pull requests through one /merge command

Scaffolds, clones, and deploys config-driven NextJS landing pages built on shadcn/ui and Tailwind v4

Sets up production monitoring for NestJS and Next.js

Designs multi-agent architectures that split work across agents to isolate context

Guides NestJS 12 APIs with Prisma and Postgres

Designs BullMQ queues, processors, and async job workflows in NestJS for video/image processing

Writes NestJS Vitest tests — testing modules, provider mocks, service/controller specs, Supertest e2e

Writes Nextra documentation sites with MDX

Reviews comments on a diff, deletes narration, fixes accepted findings, and offers encodings for claimed constraints

Audits a private repo once before it goes public

Coordinates all existing project sessions from one session, collects reports, routes questions once, and follows issues and PRs through delivery

Designs TypeScript monorepo packages

Optimizes backend, database, and delivery performance

Initializes Playwright end-to-end testing for Next.js/React projects

Runs the final refinement pass on a finished feature

Operate a production Postgres database

Triages a PR's review threads into a severity-tagged digest, or in address mode implements the fixes and drafts per-thread replies for approval

Routes /prd prepare to complete issue preparation and retains focused requirements, planning, draft lint, intake, and discovery modes

Validates draft requirements and blocks execution until the issue holds a complete, current implementation contract

Publishes prepared requirements and plans as one GitHub feature issue, with child issues only when justified

Writes repository-grounded requirements with full scope and observable EARS acceptance criteria

Audits an app's production readiness from code, CI, config, migrations, observability, and deploy paths, returning a ship/block verdict

Audits a project board's fields, columns, and scope, and applies requested setup, copy, or normalization on GitHub or Jira boards

Picks the project initialization route and orchestrates setup, preferring npx @shipshitdev/v0 for new Shipshit.dev repos

Applies prompt engineering patterns

Builds a throwaway prototype to answer one design question about state logic or UI look

Playbook orchestrator for verified, unslopped engineering work

Runs an interactive localhost QA session that accepts issues, screenshots, routes, and browser evidence; watches app, API, browser, and network errors; reproduces and fixes each defect; verifies the result; and creates one local commit per fix

Runs a multi-phase verification pass on finished AI agent work to catch bugs, missed requirements, and bad assumptions

Generates minimal HTML pages to review lists, tables, drafts, or summaries in a browser

Tones down visually aggressive or overstimulating designs while preserving quality

Diagnoses slow React components and applies render-time fixes

Applies 41 performance rules for React Hook Form

Guides React Native component building

Applies modern React patterns — hooks, composition, performance, TypeScript

Guides architectural React refactoring

Covers React Testing Library for component and hook tests

Rebuilds recent working context from chat history, live state, and the shared record into a tight current-state brief

Evaluates incoming review feedback technically before changing code

Implements Redis caching, rate limiting, sessions, locks, and pub/sub for Next.js/NestJS/Prisma apps

Refactors code safely behind tests without changing behavior

Cuts a release from a green trunk

Investigates a question against primary sources and writes one Markdown note citing each claim

Reviews a finished agent session for friction and proposes environment fixes (checks, pointers, reviewer rules, steering cuts, skill edits), most severe first

Routes /review for diffs, one/all PRs, commit windows, retros, or Grok second opinions

Turns the ICP in .agents/memory/icp.md into a revenue-ranked roadmap, finding gaps that block landing, retaining, or expanding customers

Turn a revenue-ranked roadmap into tracked GitHub Milestones with due dates, assign issues to them, and report burndown

Detects user preferences and workflow rules in conversation and records them in .agents/memory/captured-rules.md for promotion

Generates incremental code modules inside an existing repo following local patterns

Runs a scoped security audit of a web app or API

Applies OWASP Top 10 security practices to React, Next.js and NestJS code

Writes an Agent skills routing block in CLAUDE.md/AGENTS.md plus docs/agents/ recording issue tracker, kanban labels and domain docs

Configures canonical Shipshit Pstack adapters for selected harnesses while preserving existing role choices

Applies shadcn/ui best practices (Radix, forms, tables, theming) and sets up shadcn/ui on Tailwind v4 (init, components.json, cn, v3 migration)

Plan the UX and UI for a feature before writing code

Keeps a TSV decision log (what, why, evidence, result) for long or unattended work

Turns workflows and hard-won knowledge from a conversation into reusable SKILL.md files

Measures whether agents follow a skill, rule or agent definition by running scenarios and comparing action timelines to it

Guides creating or updating a skill with specialized knowledge, workflows or tool integrations

Router behind /skill

Searches local, marketplace, GitHub, package and web sources for existing skills or implementations

Modernizes a stack: upgrades outdated dependencies, removes dead or deprecated packages, migrates stale framework patterns, verifying latest versions first

Validates Biome 2.3+, Bun workspace, Next.js 16 and Tailwind v4 config and flags deprecated patterns

Summarizes personal git activity by default, all authors' integrated changes with /standup all 24, and an opt-in merged-history audit with /standup all 24 audit

Implements Stripe payments, subscriptions, webhooks and customer management in Next.js and NestJS. Use for checkout, payment intents, billing, Stripe Connect

Reviews a PR diff for structure and maintainability

Fans out N parallel workers, drains them, and returns one aggregated report

Designs filtering UX for data tables

Applies Tailwind CSS v4 best practices when writing, reviewing or refactoring styles

Drives feature work and bug fixes test-first with red-green-refactor and vertical slices

Explains a change, subsystem or concept in one plain, diagram-built account so the person truly understands it; changes nothing

Ranks a codebase's tech debt into a register scored by interest over principal, each item tied to file:line

Applies a layered technical-writing standard (Diataxis, Google style, STE) to docs, RFCs, READMEs, PR descriptions and commit messages

Router behind /test

Runs tests at the right scope (changed, focused, full, type-check, e2e) in Bun repos and reports failures with evidence; repairs only if authorized

Installs Vitest and GitHub Actions CI with Bun and 80% coverage for Next.js, NestJS, or React

Sets testing strategy: test level, coverage targets, test data, flake diagnosis, plus regression tests for AI blind spots

Styles slides, docs and HTML pages with one of 10 preset color/font themes or a generated custom theme

Designs agent-facing tools: descriptions, schemas, responses, errors, naming, and consolidating overlapping tool sets

Guides Turborepo monorepos: turbo.json tasks, dependsOn, caching and remote cache, --filter/--affected, internal packages

Solves TypeScript/JS problems and refactors TS for type safety

Deploy a Next.js app to Vercel and manage its environment

Blocks done, fixed or passing claims until the verification command has run fresh and its full output was read

Re-pitch the last message in plain English using the project's glossary vocabulary

Charts a foggy effort too big for one session as a map issue of decision tickets, then resolves one ticket per session until the way is clear

Coordinates a weekly engineering review of board accuracy, recent code changes, operational health and scoped cleanup

Investigates why code is shaped as it is

Generates an interactive bash wizard guiding a human through manual steps

Orchestrates a full-stack monorepo performance audit across frontend, backend, database, extensions and shared packages

Creates an isolated git worktree outside the repository, in the agent's worktree home, off the right base branch

Resolves implementation decisions into an exact, current execution plan on the same issue