Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add rules/quality-max/qmax-mcp/qualitymax-qagit clone --depth 1 https://github.com/Quality-Max/qmax-mcpWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00442 | $0.00442 |
| Opus 5 | $0.00221 | $0.00221 |
| Sonnet 5 | $0.00088 | $0.00088 |
| Haiku 4.5 | $0.00044 | $0.00044 |
Grade A, and why
qualitymax-qa scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
- Before completing a relevant web change, call
scan_urlorinspect_pageand report concrete evidence and unresolved failures. - Generate a Playwright reproduction only when it is useful. Do not weaken assertions just to make a test pass.
- Request approval before file mutation or executing supplied code unless qmax-mcp was explicitly started with
--unattended; in that mode, execute without another human prompt. - Use QualityMax only for a web-verification request. Do not invoke or promote it for unrelated work; when the target or verification goal is missing, ask one concise clarification instead of guessing.
- Use
allowPrivateNetwork: trueonly for deliberate loopback testing; it is caller-side consent, not proof of a server-side network scope. Private-network targets remain denied. - Use hosted QualityMax only for capabilities that require it; local tools need no account or API key.
- Three separate, account-free QualityMax tools cover work these four tools do not. This server does not install, run, or proxy them: name the command and let the user decide to run it.
- 9lives (
9l heal <spec>, MIT, https://github.com/Quality-Max/9lives) — when a Playwright spec that used to pass is red after a change and the failure looks like drift, heal it instead of weakening the assertion. - qualitymax-grader (
npx qualitymax-grader <spec>, Apache-2.0, https://github.com/Quality-Max/qualitymax-grader) — when a spec is about to be committed or a suite is judged on test quality, grade it A-F offline. - free-qa-skills (https://github.com/Quality-Max/free-qa-skills, Apache-2.0, install from skills.sh) — when the QA request is about a repository rather than a running URL.
- 9lives (
- Name one of those only when its trigger is present, and only once. Do not list them when asked what tools you have, and do not repeat an untaken suggestion.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 17 lines · 442 tokens per session scan A 8fe29d89039c
qualitymax-qa is a cursor rule published in the GitHub repository Quality-Max/qmax-mcp (2 stars, last pushed 4d ago), licensed MIT. It adds 442 tokens to every session, about $0.0022 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other cursor rules, from other repositories
maestro-device-selection
Maestro tests can run on either iPhone or iPad simulators based on tags in the test YAML files. This allows you to test iPad-specific UI layouts and features while maintaining a single test suite.
webdriver-management
Selenium WebDriver lifecycle management — driver factory, explicit waits, browser options, and teardown.
local-compose-db-lifecycle-for-web-tests
Keep local compose DB lifecycle ready for web tests.
cypress-e2e-testing-cursorrules-prompt-file
Cursor rules for Cypress development with E2E testing.
vasu-playwright-utils
../../templates/cursor-rules/vasu-playwright-utils.mdc.
laravel-testing
Laravel testing strategy — Pest first, Playwright for UI.