Run a performance verification against a threshold or SLO -- a benchmark, a load test, or a stress test -- and report measured against target. Use this skill when you need to know whether something is fast enough, and by how much. Read-only; it resolves nothing, and findings hand off to `/aid-fix`. A thin kind-sibling of `/aid-test`, which defines its full behavior.
Installs into .claude/skills of the current project.
Are you the author of Aid Test Performance?
Add the live security badge to your README. It updates with every re-scan.
[](https://www.skillsdirectory.com/skills/andrevianna-aid-test-performance)
---
name: aid-test-performance
description: >
Run a performance verification against a threshold or SLO -- a benchmark, a load test, or
a stress test -- and report measured against target. Use this skill when you need to know
whether something is fast enough, and by how much. Read-only; it resolves nothing, and
findings hand off to `/aid-fix`. A thin kind-sibling of `/aid-test`, which defines its
full behavior.
allowed-tools: Read Glob Grep Bash Shell Terminal Write Edit Agent
argument-hint: "<target + threshold> -- the hot path/endpoint and the SLO to measure against"
---
# Test Performance (performance kind-sibling of /aid-test)
`/aid-test-performance` is a thin **kind-sibling** of **`/aid-test`**
(`.agents/skills/aid-test/SKILL.md`) -- not an alias: its own catalog row
(`alias_of: null`, `{verb: test, artifact: performance}`), `repurpose: true`
(hand-authored). It carries **no logic of its own.**
Execute `.agents/skills/aid-test/SKILL.md` exactly as written, with the verification
**kind bound to performance** (capture the workload profile -- concurrency/rate/data volume
-- the threshold/SLO, and the environment; the result must be reproducible). Report
measured-vs-threshold with the workload + environment noted. Substitute only the invocation
name (`/aid-test-performance`) in any printed usage example. Findings route to `/aid-fix`.
