ddd13e6422
Group G — markdown tech-debt cleanup (post-audit 2026-05-02).
- 36 SKILL.md files: added "## When to use" section. Was missing across the
catalog; orchestrator routing by keyword could not auto-dispatch.
- 20 code-implementer agent .md files: added Output Footer block prescribing
RULE 0.16 STATUS-TRUTH MARKER schema in agent's final report. Previously only
code-implementer-rust.md had it; other 27 language/role variants were silent
about the marker, breaking RULE 0.16 §3 status-truth aggregation for non-Rust
batches.
- skills/site-create/: added phase-5-preview.md and phase-6-deploy.md skeleton
files. SKILL.md table-of-contents referenced 7 phases; only 5 existed on disk.
- skills/{ai-animation,rag-pipeline}/skill.md: added migration banner comment
noting they should be SKILL.md (canonical filename). Case-rename via git is a
separate orchestrator task (macOS APFS is case-insensitive; Linux deploy needs
explicit rename).
- 3 deprecated skills (site-builder, competitor-analysis, design-inspiration):
added concrete removed-after dates (was vague "before v2").
- docs/CONVERGENCE-PLAN.md:129: TBD on _blocks/evidence-grading.md duplicate
resolved (file exists, not duplicated).
- docs/DNA-INDEX.md: count edits made then overwritten by auto-encyclopedia-refresh
hook during agent run. The .kei-registry-ignore files in test fixtures (Group F)
are the structural fix; kei-registry walker implementation is the follow-up.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
57 lines
2 KiB
Markdown
57 lines
2 KiB
Markdown
---
|
|
name: perf-audit
|
|
description: Use when auditing performance — baseline, profile, identify top 3 bottlenecks, fix, remeasure
|
|
arguments:
|
|
- name: target
|
|
description: "What to audit: endpoint, page, function, or 'full'"
|
|
required: true
|
|
---
|
|
|
|
# Performance Audit Workflow
|
|
|
|
## When to use
|
|
|
|
- Auditing an API endpoint, page, or function for performance regressions (baseline → profile → fix → remeasure).
|
|
- Identifying the top 3 bottlenecks (DB N+1, sequential network calls, CPU hotspots, bundle bloat).
|
|
- Before shipping a feature that touches a known hot path and needs measured before/after numbers.
|
|
|
|
## Step 1: Establish Baseline
|
|
- Measure current performance:
|
|
- API: response time (p50, p95, p99), throughput
|
|
- Frontend: LCP, FID, CLS, bundle size
|
|
- Function: execution time, memory usage
|
|
- Record numbers BEFORE any changes
|
|
- Use project's existing tools or:
|
|
- Python: `time`, `cProfile`, `memory_profiler`
|
|
- JS/TS: `performance.now()`, Lighthouse, `webpack-bundle-analyzer`
|
|
- API: `curl -w @-` timing, `ab`, `wrk`
|
|
|
|
## Step 2: Profile
|
|
- Identify WHERE time is spent:
|
|
- Database queries (N+1, missing indexes, full scans)
|
|
- Network calls (sequential vs parallel, caching)
|
|
- CPU (algorithmic complexity, unnecessary computation)
|
|
- Memory (leaks, large allocations, unnecessary copies)
|
|
- I/O (file reads, disk writes)
|
|
|
|
## Step 3: Identify Top 3 Bottlenecks
|
|
- Rank by impact (% of total time)
|
|
- Focus on top 3 — don't optimize everything
|
|
- For each: document what, why slow, potential fix
|
|
|
|
## Step 4: Checkpoint
|
|
- `checkpoint: before perf-audit $target`
|
|
|
|
## Step 5: Fix (One at a Time)
|
|
- Fix #1 bottleneck → measure → confirm improvement
|
|
- Fix #2 bottleneck → measure → confirm improvement
|
|
- Fix #3 bottleneck → measure → confirm improvement
|
|
- After each fix: run tests — no regressions
|
|
|
|
## Step 6: Final Measurement
|
|
- Re-run baseline measurements
|
|
- Compare before/after
|
|
- Report: metric, before, after, improvement %
|
|
|
|
## Step 7: Commit
|
|
- `perf: optimize $target — <summary of improvements>`
|