Agent skill
benchmark
Performance benchmarking. Baseline page load times, Core Web Vitals, and resource sizes. Compare before/after on every PR. Catch bundle size regressions before they ship.
Install this agent skill to your Project
npx add-skill https://github.com/DenchHQ/DenchClaw/tree/main/skills/gstack/benchmark
SKILL.md
Benchmark — Measure Performance
Establish baselines and catch regressions.
Related skills: qa | review | ship
What to Measure
Core Web Vitals
- LCP (Largest Contentful Paint) — Target: < 2.5s
- FID (First Input Delay) — Target: < 100ms
- CLS (Cumulative Layout Shift) — Target: < 0.1
Page Load
- TTFB (Time to First Byte)
- DOMContentLoaded
- Load event
- Full page load (all resources)
Bundle Size
- JavaScript bundle size (compressed + uncompressed)
- CSS bundle size
- Total transfer size
- Number of requests
Resource Analysis
- Largest assets by size
- Uncompressed assets that should be compressed
- Unused JavaScript (tree-shakeable)
- Image optimization opportunities
Workflow
1. Baseline (before changes)
git stash # or checkout base branch
# Run measurements
git stash pop # or checkout feature branch
2. Current (after changes)
Run the same measurements on the current branch.
3. Compare
PERFORMANCE REPORT
═══════════════════════════════════════
Before After Delta
LCP: 2.1s 2.3s +200ms ⚠️
CLS: 0.02 0.01 -0.01 ✓
JS Bundle: 245KB 312KB +67KB ⚠️
CSS Bundle: 42KB 43KB +1KB ✓
Requests: 23 25 +2
Transfer: 890KB 1.1MB +210KB ⚠️
═══════════════════════════════════════
4. Flag Regressions
- JS bundle size increase > 10KB → warning
- LCP increase > 200ms → warning
- CLS increase > 0.05 → critical
- New unoptimized images → warning
Important Rules
- Always compare against the base branch, not an absolute value
- Measure multiple times and take the median to reduce noise
- Flag regressions but don't block — some features legitimately increase size
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
DenchHQ/DenchClaw
dench-integrations
Connected app integration recipes for Dench Integrations (Gmail, Slack, GitHub, Notion, Google Calendar, Linear, Stripe, YouTube, and 500+ more)
DenchHQ/DenchClaw
gstack-workflow
Structured AI-assisted development workflow with specialist roles — Think, Plan, Build, Review, Test, Ship, Reflect. Adapted from garrytan/gstack (MIT).
DenchHQ/DenchClaw
plan-ceo-review
CEO/founder-mode plan review. Rethink the problem, find the 10-star product, challenge premises. Four modes: SCOPE EXPANSION, SELECTIVE EXPANSION, HOLD SCOPE, SCOPE REDUCTION. Use when reviewing strategy, questioning scope, or before engineering review.
DenchHQ/DenchClaw
investigate
Systematic root-cause debugging. Four phases: investigate, analyze, hypothesize, implement. Iron Law: no fixes without root cause investigation first. Use when debugging errors, unexpected behavior, or troubleshooting.
DenchHQ/DenchClaw
design-consultation
Build a complete design system from scratch. Research the landscape, propose safe choices AND creative risks, generate DESIGN.md. Use when creating a new design system or establishing design foundations for a project.
DenchHQ/DenchClaw
land-and-deploy
Merge the PR, wait for CI and deploy, verify production health. Takes over after ship. One command from "approved" to "verified in production."
Didn't find tool you were looking for?