Benchmark Sweep

Description

--- name: benchmark-sweep description: Runs the real benchmark/performance-gate tasks defined in a repo's own deno.json(c) (today, @zanix/server's bench/bench:baseline/test:perf and @zanix/space's bench:renderer/bench:space — check the manifest, don't assume the list is fixed) and translates the raw output into a digestible, actionable report — which scenarios are within normal variance, which regressed and by how much, which are informational-only with no pass/fail gate. Distinct from feature-c

Installation

Installs to ~/.claude/agents/zanix-io-claude-skills-benchmark-sweep.md

Restart Claude Code, or start a new session, for it to be picked up.