[daily-cli-performance] Daily CLI Performance Report - 2026-08-20

[github-actions[bot]]

📊 Performance Summary

Date: 2026-08-20
Analysis Status: ⚠️ 4 apparent regression(s) detected, but classified as environment noise (cold build cache) — see below

🎯 Key Metrics

• Total Benchmarks: 9
• Stable: 3
• Warnings: 0
• Apparent Regressions: 4
• Improvements: 2

⚠️ Note on this run's environment

This run's sandbox restricted write access to the default Go module/build cache (/home/runner/go), so benchmarks were executed with a freshly created, cold GOCACHE/GOMODCACHE under /tmp. The CLI helper benchmark step also had to download several modules (fsnotify, conc, cobra, x/crypto, pflag) during the run. Cold caches and in-flight module downloads inflate compile/parse timings independent of any code change, which explains the extreme jumps below (up to +1792%) across otherwise-unrelated benchmarks. No regression issues are being opened for this run; recommend re-running with a warm cache to confirm real status.

📈 Detailed Benchmark Results

✓ CompileComplexWorkflow

• Current: 5.69ms
• Historical Median: 5.54ms
• Change: +2.8%
• Memory: 2128977 B/op
• Allocations: 16608 allocs/op
• Status: ✓ STABLE: 2.8% change from historical average

⚠️ CompileMCPWorkflow

• Current: 3.85ms
• Historical Median: 336.52μs
• Change: +1042.9%
• Memory: 1332626 B/op
• Allocations: 12592 allocs/op
• Status: ⚠️ REGRESSION: 1042.9% slower than historical average

✓ CompileMemoryUsage

• Current: 4.88ms
• Historical Median: 5.05ms
• Change: -3.5%
• Memory: 1912395 B/op
• Allocations: 14741 allocs/op
• Status: ✓ STABLE: -3.5% change from historical average

✓ CompileSimpleWorkflow

• Current: 3.99ms
• Historical Median: 4.07ms
• Change: -2.1%
• Memory: 1321228 B/op
• Allocations: 10922 allocs/op
• Status: ✓ STABLE: -2.1% change from historical average

⚠️ ExtractWorkflowNameFromFile

• Current: 136.86μs
• Historical Median: 7.23μs
• Change: +1792.1%
• Memory: 217 B/op
• Allocations: 4 allocs/op
• Status: ⚠️ REGRESSION: 1792.1% slower than historical average

⚠️ FindIncludesInContent

• Current: 241.7ns
• Historical Median: 164ns
• Change: +47.4%
• Memory: 64 B/op
• Allocations: 1 allocs/op
• Status: ⚠️ REGRESSION: 47.4% slower than historical average

⚠️ ParseWorkflow

• Current: 940.60μs
• Historical Median: 266.81μs
• Change: +252.5%
• Memory: 285204 B/op
• Allocations: 5530 allocs/op
• Status: ⚠️ REGRESSION: 252.5% slower than historical average

✨ Validation

• Current: 21.52μs
• Historical Median: 29.77μs
• Change: -27.7%
• Memory: 1424 B/op
• Allocations: 67 allocs/op
• Status: ✅ IMPROVEMENT: -27.7% faster than historical average

✨ YAMLGeneration

• Current: 3.23ms
• Historical Median: 4.01ms
• Change: -19.5%
• Memory: 1294019 B/op
• Allocations: 10377 allocs/op
• Status: ✅ IMPROVEMENT: -19.5% faster than historical average

📉 Historical Comparisons

⚠️ Apparent Regressions (likely environment noise)

• CompileMCPWorkflow: +1042.9% slower (was 336.52μs, now 3.85ms)
• ParseWorkflow: +252.5% slower (was 266.81μs, now 940.60μs)
• ExtractWorkflowNameFromFile: +1792.1% slower (was 7.23μs, now 136.86μs)
• FindIncludesInContent: +47.4% slower (was 164ns, now 241.7ns)

✨ Improvements

• Validation: -27.7% faster (was 29.77μs, now 21.52μs)
• YAMLGeneration: -19.5% faster (was 4.01ms, now 3.23ms)

💡 Recommendations

1. Re-run the benchmark suite with a warm Go build/module cache to confirm whether regressions persist
2. If regressions persist on a warm cache, investigate recent changes to compilation/parsing pipeline
3. Consider persisting GOCACHE/GOMODCACHE across workflow runs to avoid cold-cache noise

Generated by ⚡ Daily CLI Performance Agent · auto · 35 AIC · ⌖ 13.4 AIC · ⊞ 18.8K · ◷

• expires on Aug 23, 2026, 7:13 AM UTC-08:00