claude-configuration

Expert guidance for creating, building, and using Claude Code agents and the Task tool. Use when working with agents, setting up agent configurations, understanding how agents work, or using the Task tool to launch specialized agents.

Complete SaaS ideation to implementation workflow - from idea discovery to task breakdown

Run structured APEX implementation (Analyze, Plan, Execute, eXamine) with parallel agents, tests, adversarial review, and proof-backed verification. Use for features, bug fixes, or code changes that need disciplined execution and real-flow evidence.

Deep iterative research using progressive flow psychology (diverge-analyze-converge-execute) with parallel agents, skeptical analysis, and multi-perspective synthesis. Use for thorough topic exploration, decision-making research, or when you need battle-tested conclusions.

Automated CI/CD pipeline fixer - watches CI, fixes errors locally, commits, and loops until green. Use when CI is failing and you want to automatically fix and verify changes.

Improve unclear UX copy, error messages, microcopy, labels, and instructions. Makes interfaces easier to understand and use.

Clean Code rules for TypeScript, React, and Next.js. Apply when writing, refactoring, or reviewing code - naming, small functions, no any, early returns, typed errors, plus tech-specific rules for React 19, Next.js 15/16, Zustand v5, TanStack Query v5.

Commit, push, then monitor Vercel deployment and GitHub Actions until green. Auto-fixes failures and retries. Use when committing code that needs deployment verification, or when you want to ship and verify in one command.

Quick commit and push with minimal, clean messages

Write, rewrite, and sharpen high-converting copy. Use when the user asks for emails, newsletters, headlines, hooks, landing-page or sales-page copy, ad scripts, VSL, CTAs, offers, or persuasion and conversion rewriting.

Run consensus code review with parallel specialist agents. Use when the user wants Crab Method review, high-confidence fixes, or audits for clean code, security, logic, performance, or accessibility.

Create and push PR with auto-generated title and description

Systematic error debugging with analysis, solution discovery, and verification

Browser automation with persistent page state using dev-browser CLI. Use when user mentions 'browse', 'open website', 'click on', 'fill form', 'take screenshot', 'scrape', 'automate browser', 'test website', 'log into', 'navigate to', 'web page', 'dev-browser', or any browser interaction task.

Set up per-worktree environments for Claude Code, Cursor, or Codex. Use for worktree-ready repos, IDE environment config, worktree-up/down scripts, or dev.sh wiring.

Search the web with the local Exa CLI. Use for Exa-powered research, recent sources, domain-filtered search, similar-page discovery, URL content extraction, or cited web answers.

Fast exploration of codebase, documentation, and web. Use when understanding code structure, researching libraries, or gathering context for implementation.

Retrieve current technical docs, API references, config options, and code examples. Use before answering or coding against libraries, frameworks, SDKs, CLIs, cloud services, or APIs.

Fix all ESLint and TypeScript errors with parallel processing using snipper agents

Fix grammar and spelling errors in one or multiple files while preserving formatting

Fetch PR review comments and implement all requested changes

A relentless interview to sharpen a plan or design.

Build and audit marketing strategy and go-to-market. Use when the user asks for positioning, offers, funnels, launch plans, ads, lead magnets, audience research, brand messaging, or campaign strategy.

Intelligently merge branches with context-aware conflict resolution

Implement one focused code change quickly with minimal exploration and targeted validation. Use for small bug fixes, single feature edits, config updates, or explicit "$oneshot" requests.

Expert prompt engineering for creating effective prompts for Claude, GPT, and other LLMs. Use when writing system prompts, user prompts, few-shot examples, or optimizing existing prompts for better performance.

Refactor code across the codebase using parallel Snipper agents. Use when user wants to rename methods, update patterns, fix code smells, or apply consistent changes across multiple files.

Run deep multi-agent code or PR review. Use when the user asks to review code, audit a PR, check security or quality, or find high-impact issues beyond nitpicks.

Create, edit, and maintain AGENTS.md and .agents/rules/ — tech stack, commands, universal rules, rule index, and rule files. Use to add, modify, restructure, or optimize project rules, conventions, and constraints.

Create or edit Claude, Codex, and Cursor skills/rules. Use for SKILL.md, .cursor/rules, AGENTS.md, skill prompts, frontmatter, references, scripts, and discovery rules.

Create multi-step workflow skills with progressive step loading, state management, and interactive decisions. Use when building complex skills like APEX that need structured phases.

Run an extremely strict maintainability review for abstraction quality, giant files, and spaghetti-condition growth. Use for a thermo-nuclear code quality review, thermonuclear review, deep code quality audit, or especially harsh maintainability review.

Deep thinking mode - approach problems like a craftsman, obsess over details, and create elegant solutions

Create Claude-style local HTML artifacts under global ~/.agents/artifacts. Use for artifact requests, variations/options croquis, feature plans, thinking docs, prototypes, visualizations, dashboards, diagrams, or substantial reusable content. Pair with use-style.

Prove that a feature, fix, application, workflow, API, or CLI actually works through current runtime evidence. Use when the user asks to verify, prove, demonstrate, confirm, test end to end, browser-test, capture screenshots, validate a fix in the real app, or refuses an unproven completion claim.

Run an authorized black-box, gray-box, or white-box pentest against a live website — recon, auth bypass, privilege escalation, IDOR, injection, XSS — with mandatory PoC verification and a severity-rated report. Use when asked to pentest, attack, or security-test a site.

Sync NEW files from this repository to ~/.claude

Sync files from ~/.claude to claude-code-config repository

Auto-update the project changelog from git history. Analyzes commits, categorizes changes, and writes human-readable monthly changelog files.