What problem does it solve? Measuring how fast a Node.js function actually runs requires careful setup: realistic arguments, mocked dependencies, and statistically meaningful repetition. This Skill automates that process by deriving benchmark inputs from real call sites and tests in your codebase, running the benchmark through the ns-benchmark MCP server, and producing a structured performance report. ## Core Features & Use Cases - Evidence-based argument generation: Searches your workspace for real invocation sites and tests to build realistic args and argSetupCode mocks for complex dependencies like databases, file systems, and event emitters. - Configurable benchmark engine: Controls repeatSuite, minSamples, minTime, and maxTime, with a mandatory user confirmation step before any benchmark runs. - Structured reporting: Produces a markdown report with ops/sec, iterations, histogram min/max, variance assessment, and a full tool execution log, optionally saved to .nsolid/assets/. - Use Case: You suspect a utility function is a hot-path bottleneck. Point the Skill at the function, confirm the proposed arguments derived from its unit tests, and receive a table showing ops/sec and variance diagnostics suggesting whether JIT tier transitions or GC pauses are affecting consistency. ## Quick Start Ask the AI to benchmark a specific function in your workspace, for example: benchmark the parseQuery function in src/utils.js and show me its ops/sec.
