Use this when reviewing a completed feature for performance risks across API, database, and frontend paths. Typically run after implementation and verification, before finalization. Do not use for planning, implementation, debugging, or testing.
Installation
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.
Use this when reviewing a completed feature for performance risks across API, database, and frontend paths. Typically run after implementation and verification, before finalization. Do not use for planning, implementation, debugging, or testing.
Always read
agents.md
architecture.md
repo_map.md
External documentation rule
If performance review depends on framework, library, ORM, or SDK behavior, use Context7 to consult current official documentation before making assumptions.
Prefer repository code as the source of truth for local behavior and architecture.
If repo code and external docs conflict, call out the conflict explicitly and favor the repo for existing implementation intent.
When to use
after implementation is complete and verified
when changes touch performance-sensitive paths (API, DB, rendering)
Inputs
feature_slug
docs/plans/<feature_slug>_plan.md
modified files (git diff)
relevant API, DB, frontend, or rendering logic
Output
concise performance summary
concrete risks with impact and locations
practical fix recommendations
no-action notes where optimization is unnecessary
Before reviewing
read docs/plans/<feature_slug>_plan.md
review modified files (git diff)
inspect API, DB, and frontend/rendering paths
do not modify code; this step is read-only
Rules
avoid premature optimization; focus on real bottlenecks
do NOT invent issues; base findings on code and patterns observed
prefer simple, low-risk improvements
treat code as source of truth over documentation
be concise and structured
Checks
Evaluate for:
N+1 queries / inefficient DB access
inefficient loops or algorithms (time/space)
large payloads / over-fetching
unnecessary re-renders or state churn
repeated computation / lack of memoization
duplicate or redundant API calls
blocking operations on request/response path
missing pagination / indexing opportunities
unbounded concurrency or missing timeouts
Output Format
Performance Summary
[concise]
Risks
[issue]
Impact: low | medium | high
File: path
Why: short
Fix: short
No-action Notes
[bullets]
Recommendations
[bullets]
Constraints
do NOT modify any files
do NOT include unrelated commentary
keep output minimal and actionable
if context is incomplete, state assumptions briefly in No-action Notes
