Use when the user reports a bug or needs help investigating unexpected behavior. Triggers: found a bug, bug report, something's broken, this doesn't work, investigate this bug, be my debugging partner, help me debug, manual verification failed, why is this failing, unexpected behavior, regression. Investigates root cause and generates tasks — does NOT write fixes (use tdd-agent for that).
Use when the user wants to audit code quality across the repo. Triggers: code health, audit code quality, check code health, tech debt scan, find dead exports, check file lengths, find circular deps, test coverage gaps, code quality check. Spawns 6 parallel subagents to scan for issues and creates tech-debt tasks for findings. Can run standalone, or is invoked by tdd-agent (changed files) and pm-agent (full scan).
Use when the user wants to scan the codebase and fill in CLAUDE.md project context. Triggers: detect project, scan project, fill in CLAUDE, update project context, populate CLAUDE.md, scan codebase, what's in this project, set up project context. Run this after n2o init or whenever the project structure changes.
Multi-agent frontend review agent. Assesses UI quality using programmatic checks + LLM vision + interactive testing. Auto-fixes issues with deterministic merge rules. Triggers: frontend review, review UI, check my page, UX audit, accessibility check, visual review, review this component.
Use when the user wants to plan, scope, or break down work. Triggers: create spec, new feature spec, write spec, break down, decompose tasks, task breakdown, sprint planning, plan sprint, pm workflow, project management, planning agent, scope this feature, what should we build, let's plan, create tasks, load tasks, start a sprint. Use this for planning and orchestration — NOT for implementing code (use tdd-agent for implementation).
Consult this skill automatically when writing, reviewing, or refactoring React or Next.js code. Use as an ambient guide — check these patterns whenever creating components, data fetching hooks, server components, client components, or optimizing bundle size. Triggers: review my React code, check performance, optimize React, Next.js best practices, component performance, bundle optimization, react patterns, writing React components, refactoring React code.
Use when the user wants to implement sprint tasks using TDD. Triggers: implement task, tdd, test-driven development, task execution, implement feature, start working on task, pick next task, let's implement, build this feature, write tests first, red green refactor. This skill drives the RED → GREEN → REFACTOR → AUDIT → CODIFY → COMMIT cycle. NOT for planning (use pm-agent) or bug investigation (use bug-workflow).
Consult this skill automatically when writing or reviewing UI code. Provides 28 principle-based UX heuristic rules organized by category. Assessed by the frontend review agent. Triggers: UX review, heuristics check, UI quality, layout review, accessibility check.