Measures performance baselines, detects regressions before and after changes, and compares stack alternatives.
- What it does
- This skill defines a benchmarking workflow with four modes: page performance (Core Web Vitals and resource sizes via a browser MCP), API performance (latency percentiles and load testing), build performance (cold build, HMR, test, type-check, lint, Docker times), and before/after comparison. It stores baselines as JSON under .ecc/benchmarks/ so they can be tracked in Git and shared by a team. It also describes integration with CI and with other skills for post-deploy monitoring and pre-ship checklists.
- When to use it
- Use it when measuring the performance impact of a pull request, setting a project performance baseline, investigating reports that something feels slow, verifying performance targets before launch, or comparing a stack against alternatives.
- Requirements
- Instructions only; no scripts are shipped. A browser MCP is referenced for page metrics, and API, build, and comparison modes assume the relevant endpoints, build tooling, and Git-tracked repository are available.