Benchmark Report Skill for ChatGPT: Compare Throughput, P95 and Concurrency
Structure completed performance test results with deployment configuration, load levels, success rates, and latency percentiles. Keep missing data and incompatible conditions visible.
中文Install in ChatGPT
Requires a ChatGPT workspace where you can create skills. This is a text adaptation of the pinned GitHub source.
- Review the inputs and outputs in the pinned .claude/skills/benchmark-report/SKILL.md.
- In ChatGPT, open Plugins, Skills, Create, and Create with editor.
- Write the skill rules using the workflow described here. Provide completed result excerpts or JSON, deployment configuration, and metric definitions.
- Save the skill and check the fictional example in a new chat.
Example task
Fictional search results show P95 of 120 ms and throughput of 8/s at concurrency 1, versus 260 ms and 42/s at 10. Registration data is absent. Draft a summary and limitations.What to check
Preserve both load settings, note higher throughput and higher P95 at ten, and mark registration unavailable. Record which corpus and environment details need confirmation.
Original SKILL.md on GitHub · Updated
Frequently asked question
Does higher P95 establish a capacity bottleneck?
It shows an increase in tail latency. Use error rates, resource observations, sample size, and additional comparable loads to investigate a cause.