Web3-Claude

Prepares commits and generates PR descriptions from commit and file analysis. Auto-detects existing PRs to offer create or update commands. Writes to /tmp/pr-body-<repo>.md and gives user the gh command. Invoke when work is ready for PR.

Top-level orchestration agent for all implementation work. Analyzes tasks, selects the right agents, coordinates multi-domain work, and ensures knowledge files stay in sync. Consult this agent before starting any non-trivial implementation. Manages ui-designer and web3-implementer as primary subagent trees, plus shared specialists.

Adversarial person vetting agent. Investigates people before engaging, hunting identity fraud, hacked accounts, and social engineering scams. Default stance is guilty until proven innocent. Primary focus on LinkedIn contacts; also supports Twitter, Discord, Telegram, and conference contacts.

Interactive adversarial DeFi protocol research. Chat about protocol risks, investigate teams, analyze on-chain data, and get integration recommendations. Default stance is guilty until proven innocent. Invoke with /protocol-risk-analyst.

Creates new custom React hooks following project conventions. Scaffolds TypeScript types, JSDoc documentation, and proper error handling in src/hooks/. References hook-patterns.md for layer-appropriate templates.

Syncs .claude knowledge files with the codebase. Detects stale reference files (hook catalogs, type indexes, schema references, theme docs) and regenerates them from source. Run after adding hooks, types, components, or schema changes.

Creates new Claude Code skills in the .claude/skills/ directory. Specializes in designing well-structured, effective skills through thorough requirements gathering. Use when the user wants to create a new skill, command, agent, or automation.

Analyzes MUI theme compliance across all components. Finds hardcoded colors, font sizes, font weights, and rem units. Generates structured violation report with exact fixes, then offers to apply fixes automatically. Use for theme auditing or when suspecting hardcoded styles.

Safety monitor that detects dangerous patterns during agent operations. Catches destructive commands (rm -rf, force push), malicious packages, and secrets exposure. Blocks and alerts on dangerous actions. Invoke directly to review plans or recent actions.

Creates new React components following project conventions. Scaffolds TypeScript types, MUI styling patterns, and named exports in src/components/. References Common component library to avoid duplication.

Ralph Loop plugin manager. Provides start, cancel, status, and help commands for autonomous task loops. Enforces safety guardrails (sandbox, deny rules, PR-only, max-iterations). Use /ralph-loop:start to begin, /ralph-loop:cancel to stop.

Comprehensive application verification specialist. Runs static analysis (typecheck, lint, prettier), build verification, code quality review, and security checks. Produces structured pass/fail report with recommendations.

Lightweight project verification. Runs typecheck, lint, prettier (auto-fix), and build. Reports pass/fail status and offers to auto-fix lint issues. Use for quick verification before commits or after changes.