Build performance optimized

Severity: Recommended

Check automatically:

1. Check for caching in workflows:

   # Look for cache action usage
   grep -riE "actions/cache|cache:|restore-keys" .github/workflows/*.yml 2>/dev/null
   

2. Check for dependency caching:

   # Node/pnpm cache patterns
   grep -A10 "actions/cache\|pnpm/action-setup\|actions/setup-node" .github/workflows/*.yml 2>/dev/null | grep -iE "cache|pnpm-store|node_modules|\.npm"
   

3. Check for Docker layer caching:

   # Docker buildx cache
   grep -riE "cache-from|cache-to|buildx.*cache|docker/build-push-action" .github/workflows/*.yml 2>/dev/null
   

4. Measure actual build times:

   # Get recent workflow run durations (deployment workflows)
   gh run list --limit 20 --json workflowName,createdAt,updatedAt --jq '[.[] | select(.workflowName | test("deploy|release|build"; "i"))] | .[] | {name: .workflowName, duration_minutes: ((((.updatedAt | fromdateiso8601) - (.createdAt | fromdateiso8601)) / 60) | floor)}'
   

5. Check for parallelization:

   # Look for matrix builds or parallel jobs
   grep -riE "matrix:|needs:|concurrency:" .github/workflows/*.yml 2>/dev/null
   

6. Check cache key includes lockfile:

   # Cache keys should include lockfile hash for correctness
   grep -A5 "actions/cache" .github/workflows/*.yml 2>/dev/null | grep -E "key:.*lock|key:.*hashFiles"
   

Cross-reference with:

• DEPLOY-001 (Workflow stability)
• Section 28 (Code Architecture - build times)

Pass criteria:

• Dependency caching configured (node_modules, pnpm store, pip, etc.)
• Build times reasonable (under 5 min typical, under 10 min for complex builds)
• Cache keys include lockfile hash (prevents stale cache)
• Docker layer caching (if using Docker builds)
• Parallelization used where beneficial

Fail criteria:

• No caching configured (cold builds every time)
• Build times consistently >15 minutes
• Cache keys don't include lockfile hash (stale cache risk)
• Obvious parallelization opportunities missed

If builds slow, ask user: "Build times averaging over 10 minutes. Is this acceptable for your workflow? Long builds slow down iteration and developer feedback loops. Consider adding caching or parallelization."

Evidence to capture:

• Caching configured (yes/no, what type)
• Average build duration (minutes)
• Cache key pattern (includes lockfile hash?)
• Parallelization used (matrix, parallel jobs)
• Docker caching (if applicable)

---