Agent skill · software engineering · aiskillstore
performance-gate
Verify performance implications were considered and no obvious anti-patterns exist. Issues result in WARNINGS.
Why this skill is useful
Adds a structured performance review process with specific checks and templates that the AI wouldn't reliably generate on its own.
What it needs
About 4k tokens when loaded. Last updated 2026-08-07. 405 stars on the source repository.
What this skill does
Gate 4: Performance Review "Code that works is step one. Code that scales is step two." Purpose This gate catches performance anti-patterns before they cause problems. The focus is on obvious issues, not micro-optimizations. Gate Status PASS — No obvious performance issues WARNING — Issues found that could cause problems at scale --- Gate Questions Question 1: Scalability "What happens when there are 10,000 items? 1,000,000?" Looking for: Awareness of data growth Pagination for large datasets Efficient data structures No unnecessary loops Question 2: Query Efficiency "How many database queries does this operation make?" Looking for: No N+1 queries Bulk operations where appropriate Indexes on queried columns Awareness of query cost Question 3: Re-render Awareness (Frontend) "When this state changes, what components re-render?" Looking for: Awareness of render triggers Appropriate memoization State placement optimization No expensive computations in render --- Performance Checklist Database Operations [ ] No N+1 queries (queries inside loops) [ ] Pagination for list endpoints [ ] Indexes on frequently queried columns [ ] SELECT only needed columns (not SELECT ) Frontend Rendering [ ] Expensive computations use useMemo [ ] Event handlers use useCallback where needed [ ] Large lists use virtualization [ ] Heavy components are lazy loaded API & Network [ ] Response payloads are minimal [ ] Large data is paginated [ ] Caching headers where appropriate [ ] No redundant API calls General [ ] No nested loops (O(n²)) without justification [ ] No blocking operations [ ] Cleanup of intervals/timeouts [ ] Reasonable memory usage --- Response Templates If PASS If WARNING --- Common Issues to Check 1. The N+1 Query Problem 2. Fetching Everything 3. Expensive Render Calculations 4. Inline Functions Causing Re-renders 5. Missing Cleanup --- Socratic Performance Questions Instead of pointing out the fix, ask: 1. "How many queries does this endpoint execute for 100 users?" 2. …
How to use it
Reference it in AdaL, Claude Code, Cursor or any coding agent — nothing to install:
@skills aiskillstore/performance-gate