What problem does it solve?
This Skill eliminates the risk of manual, disruptive API performance audits that can impact production or miss critical bottlenecks in large codebases, delivering evidence-based performance insights while prioritizing safety and reproducibility.
Core Features & Use Cases
- Safe Local/Staging Benchmarking: Run conservative load tests on non-production APIs using installed tools or a bundled Python fallback, with strict safety guardrails to avoid disrupting live services.
- MADMAX Parallel Deep Scan: Automatically split large codebases into focused zones for parallel agent auditing, detecting deep bottlenecks that single-pass audits miss.
- Precision Bottleneck Detection: Identify N+1 queries, synchronous blocking, memory leaks, connection pool exhaustion, missing database indexes, and payload bloat with targeted regex and code analysis.
- Statistical Validation & Reporting: Generate concise, actionable reports with percentile latency distributions, Universal Scalability Law (USL) capacity analysis, and Mann-Whitney U test validation for optimization changes.
- Use Case: You have a 200-file Go backend with 45 API routes. Use this Skill to run a MADMAX parallel audit, identify 2 critical N+1 queries causing 900ms p99 latency, and get prioritized, approval-gated fix recommendations without modifying any code.
Quick Start
Use the safe-api-performance-auditor skill to benchmark your local backend API at http://127.0.0.1:3000 and report performance bottlenecks without making any code changes.
