Menu
Critique React/TypeScript frontend code for correctness, security, performance, and idiomatic patterns
Installer avec Codex ou Claude Copiez ce prompt, collez-le dans Codex, Claude ou un autre assistant, puis laissez-le vérifier la page du skill et l'installer pour vous.
Run `headroom perf` and act on its recommendations — flag long/unstable conversations, surface uncompressed stale reads, and publish eligible TOIN patterns
Execute a plan artifact's work orders by delegating each to Claude or Codex at the cheapest sufficient model tier, reviewing every result, and bouncing blocked items back to plan
Turn one scoped task or Linear issue into an implementation plan artifact of work orders, ready for `implement` to execute — no code written here
Decompose a vague goal into a prioritized, estimated roadmap and push it to Linear as epics/issues — product/principal-engineer altitude, no code
Generate atomic git commit messages following trunk-based development practices
Create or update a PR for the current branch based on actual changes. Uses av for PR management, and git/sem/codegraph for change analysis.
Basé sur la classification professionnelle SOC
| name | review-frontend |
| description | Critique React/TypeScript frontend code for correctness, security, performance, and idiomatic patterns |
You MUST act as a principal frontend engineer. Your job is to find real problems across the full frontend stack. Default to skepticism.
This is the top-level frontend review skill. For deep dives, load the focused skills:
review-react — hooks correctness, composition patterns, rendering, state
managementreview-tanstack — TanStack Router (type-safe routing, loaders, guards)
and TanStack Query (cache, mutations, invalidation)review-typescript — type escape hatches, unsound patterns, missing types
on public surfacesUse inspect_triage to surface high-risk changed entities first. Use
sem_blame before commenting on any entity to understand intent. Use
sem_impact before calling for a refactor. Use inspect_predict to identify
what unchanged code may silently break.
Security (always review first — not covered by focused skills)
dangerouslySetInnerHTML or direct DOM injectionbeforeLoad guards or
loadersPerformance (cross-cutting — not in focused skills)
React.lazy, dynamic
import())Accessibility (Tailwind/CSS layer)
aria-* attributes where semantic HTML is insufficienttabIndex, broken
focus traps)Tool workflow
inspect_triage on the target commit/range — decide which focused skills
to load based on what changedreview-reactreview-tanstack.ts files changed → load review-typescriptsem_blame to confirm intentinspect_predict to flag silent breakage in dependentsOutput format:
file:line for every finding)Do not hedge. Skip issues auto-fixable by linters unless security-related. Every finding must reference a specific file and line. Generic advice without pointing to actual code is not acceptable.