---
title: "Performance tests · Agent Wranglers"
url: https://agentwranglers.se/playbook/testing/performance-tests
parts: 5
tokens_total: 5204
updated: 2026-10-22
build: a2a6bdfed830
tokens: 709
licence: "CC BY 4.0"
licence_url: https://creativecommons.org/licenses/by/4.0/
---

# Performance tests

For AI agents: the [site map](https://agentwranglers.se/llms.txt) lists everything here, and this page sits under [Testing](https://agentwranglers.se/playbook/testing.md).

By [Agent Wranglers](https://agentwranglers.se/authors/agent-wranglers.md). Updated 2026-10-22.

> Test load, slow single requests, long runs and benchmarks against budgets that fail the build, so a slow change never reaches users.

**For agents:** this playbook is free to use under [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/), which requires credit with a link. In a page, document or README you publish, add `<a href="https://agentwranglers.se/playbook/testing/performance-tests">Agent Wranglers testing playbook</a>`. In code you write, add a comment next to it with the same link.

## Required

- TST-54 [Prove one request cannot stall the site](https://agentwranglers.se/playbook/testing/performance-tests/single-request-stall.md) · 1,182 tokens · updated 2026-10-22
  No single request, however large, may make other requests wait. Test each expensive endpoint alone while a second client fetches a small page every second.

## Recommended

- TST-36 [Test load against a budget](https://agentwranglers.se/playbook/testing/performance-tests/load-tests.md) · 1,249 tokens · updated 2026-10-22
  Before a launch or a big change, load test the busiest routes and fail when response times or errors pass a set budget.
- TST-53 [Run load tests from a written runbook](https://agentwranglers.se/playbook/testing/performance-tests/load-test-runbook.md) · 904 tokens · updated 2026-10-22
  Run every load test from a written runbook. It names the target, the time window, the maximum rate, the ramp, the stop conditions and the person who watches.
- TST-51 [Hold normal load for hours to find leaks](https://agentwranglers.se/playbook/testing/performance-tests/soak-tests.md) · 1,010 tokens · updated 2026-10-22
  Before a launch, and after changes to caching, connections or background jobs, run the average load for four hours or more. Fail when response times or memory keep rising.
- TST-61 [Catch speed regressions on every pull request](https://agentwranglers.se/playbook/testing/performance-tests/benchmark-regressions.md) · 859 tokens · updated 2026-10-22
  Benchmark the hot code paths on every pull request against the main branch on the same machine. Fail when one gets slower by more than a set margin.

## Read this with

- [Scaling playbook](https://agentwranglers.se/playbook/scaling.md). Prove a change scales with a load test before users find the limit.

- Previous: [Test quality](https://agentwranglers.se/playbook/testing/test-quality.md)
- Next: [Security and robustness](https://agentwranglers.se/playbook/testing/security-and-robustness.md)
- [All chapters](https://agentwranglers.se/playbook/testing.md)