api-load-tester — Claude Code skill: HTTP load testing with reports
api-load-tester — a skill for Claude Code, Codex, Cursor and other platforms that stress-tests HTTP endpoints with progressive concurrency.
MIT 260 installs ★ 440 updated 2026-10-20
npx skills add OneWave-AI/claude-skills --skill api-load-tester -a claude-codeWhat api-load-tester does
The skill opens by selecting a load tool: it checks for hey, wrk, ab, and curl in that order, then installs hey via brew or go install when none of the first three are found. A single warm-up request validates each endpoint and captures baseline latency before any load stages begin.
Stages ramp from 1 to 100 concurrent users by default — step-doubling each stage, 10 seconds per stage, 2-second pause between them. The skill prints the full test plan and waits for confirmation before running. Only GET is tested unless you supply a method, headers, and body explicitly. Default thresholds are p50 < 100 ms, p95 < 300 ms, p99 < 1000 ms; all three are configurable.
After all stages finish, the skill writes api-load-report.md to the working directory. The file includes ASCII throughput and latency charts, a scaling table, the breaking point, and specific optimisation suggestions. Auth tokens are masked in all output. Testing a production endpoint requires explicit confirmation — the skill blocks until it gets one.
What it helps with
- find the concurrency level where an API starts failing
- measure p50, p95, and p99 latency across load stages
- plot a throughput curve to find the ceiling before deployment
- classify the bottleneck and get actionable fix suggestions
- produce a shareable
api-load-report.mdwith ASCII charts for the team
Example prompts for the agent
- Run api-load-tester against
POST /api/orderswith 100 concurrent users, headerAuthorization: Bearer <token>, p95 threshold 500 ms - Load-test
GET /api/productson staging, concurrency range 1–50, 30 seconds per stage, write report toapi-load-report.md - Test all endpoints listed in
endpoints.txtand highlight any breaking points in the console summary
How to install api-load-tester
1Skills CLI (Vercel)
npx skills add OneWave-AI/claude-skills --skill api-load-tester -a claude-code2Manually: put the files in place
~/.claude/skills/api-load-tester/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/api-load-tester/
1Skills CLI (Vercel)
npx skills add OneWave-AI/claude-skills --skill api-load-tester -a codex2Manually: put the files in place
.agents/skills/api-load-tester/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add OneWave-AI/claude-skills --skill api-load-tester -a cursor2Manually: put the files in place
.cursor/skills/api-load-tester/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install OneWave-AI/claude-skills api-load-tester2Manually: put the files in place
.github/skills/api-load-tester/SKILL.md1Skills CLI (Vercel)
npx skills add OneWave-AI/claude-skills --skill api-load-tester -a gemini-cli2Platform CLI
gemini skills install https://github.com/OneWave-AI/claude-skills --path api-load-testerManually: put the files in place
.devin/skills/api-load-tester/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add OneWave-AI/claude-skills --skill api-load-tester -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about api-load-tester
How do I install api-load-tester in Claude Code?
Run npx skills add OneWave-AI/claude-skills --skill api-load-tester -a claude-code. The skill is then available at ~/.claude/skills/api-load-tester/SKILL.md.
What does the api-load-tester skill do?
It sends progressive waves of concurrent requests to your HTTP endpoints, calculates latency percentiles and error rates at each concurrency level, identifies the breaking point, and writes a full report with charts and optimisation recommendations.
Does api-load-tester require hey or wrk to be installed first?
No prior install is needed. The skill checks for hey, wrk, and ab at startup and installs hey automatically if none are present — brew install hey on macOS, go install github.com/rakyll/hey@latest on Linux with Go. If Go is unavailable it falls back to curl with background bash processes.
Source text
SKILL.md · MIT
--- name: api-load-tester description: Load tests API endpoints with progressive concurrency. Measures response times, error rates, throughput, and identifies breaking points. Generates a detailed report with latency percentiles, throughput curves, bottleneck analysis, and optimization recommendations. tools: Bash, Read, Write, Glob, Grep model: inherit --- # API Load Tester Stress-test HTTP endpoints under increasing load, identify breaking points, and produce a report with actionable recommendations. ## Contents - `references/tool-commands.md` -- tool invocations (hey/wrk/ab/curl), default concurrency stages, per-stage data to capture. - `references/metrics-interpretation.md` -- latency, throughput, error, breaking-point, and bottleneck classification. - `references/output-template.md` -- exact structure for `api-load-report.md`, including ASCII charts and scaling table. - `references/rules-and-examples.md` -- safety rules, error handling, and example invocations. ## Inputs Collect from the user. Ask before proceeding if a required input is missing. **Required**: endpoint URL(s) (with method, headers, body as needed); expected latency thresholds. Default thresholds if unspecified: p50 < 100ms, p95 < 300ms, p99 < 1000ms. **Optional**: concurrent users or range (default ramp 1 to 100); authentication; request payloads; custom headers; test duration (default 10s per stage); ramp pattern (default step ramp, doubling each stage); success criteria (default 2xx); known rate limits; environment label (prod/staging/dev). ## Workflow Follow these steps in order. 1. **Select a tool.** Check in priority order: `which hey`, `which wrk`, `which ab`, `which curl`. If none of hey/wrk/ab exist, install hey (`brew install hey` on macOS, `go install github.com/rakyll/hey@latest` on Linux with Go) or fall back to curl with bash background processes and `wait`. Verify with a single trivial request against a provided endpoint; diagnose connectivity or auth before continuing. 2. **Validate endpoints.** Send one request per endpoint with the specified method, headers, auth, and body. Confirm the status matches the success criteria and record baseline single-request latency. On failure, surface the error and ask whether to skip or fix. 3. **Design the test plan.** Build progressive concurrency stages (see `references/tool-commands.md` for the default progression), trimming or extending to the user's concurrency range. Define per-endpoint method, URL, headers, body, success codes, and timeout (default 30s). Print the plan for review before executing. 4. **Execute stages.** For each endpoint, run every concurrency stage sequentially with the selected tool, waiting 2 seconds between stages. Capture and store the per-stage metrics. See `references/tool-commands.md` for commands, request-count formula, and the metrics list. 5. **Interpret metrics.** Compute latency percentiles and profile, throughput curve and ceiling, error rates and onset, the breaking point, and the bottleneck classification. See `references/metrics-interpretation.md`. 6. **Generate the report.** Write `api-load-report.md` to the current working directory following `references/output-template.md` exactly, including ASCII throughput and latency charts. 7. **Post-report actions.** Print a 3-5 line summary to the console, state the report path, explicitly highlight any critical issues, and offer to re-run specific stages with different parameters. ## Rules Apply the safety rules, error handling, and example invocations in `references/rules-and-examples.md`. Key constraints: never load-test production without explicit confirmation, only test GET by default, mask auth tokens, respect 429 rate limits, count timeouts as failures, and never extrapolate beyond tested ranges.
© OneWave-AI · MIT · OneWave-AI/claude-skills