# QualityMax documentation > Public product documentation for QualityMax, an AI-native test automation platform. This index follows the proposed llms.txt convention. Every public documentation page is listed below in Markdown. Human-readable docs: https://docs.qualitymax.io/ ## Documentation - [Adversarial evals](https://docs.qualitymax.io/adversarial-evals.md): Run a curated library of 40 attack prompts across six categories against your AI agent and get a blocked-vs-compromised report with a block rate. - [Conversation evals](https://docs.qualitymax.io/agent-evals.md): Score your AI agent's real conversations across accuracy, helpfulness, safety, relevance, and conciseness — per-test verdicts with a configurable pass threshold. - [Agentic Eyes](https://docs.qualitymax.io/agentic-eyes.md): Agentic Eyes reviews your site from a simulated customer's perspective and reports visible observations, eligible findings, and evidence limits. - [Agentic Eyes consensus panels](https://docs.qualitymax.io/agentic-eyes-consensus.md): Run the same persona review across multiple AI models and rank findings by cross-model agreement — convergent findings are high-confidence, single-model findings land in a divergence section. - [Agentic Eyes personas](https://docs.qualitymax.io/agentic-eyes-personas.md): Built-in personas Sally (the Dreaming Planner) and the Worst Customer Ever, plus a persona library for saving custom AI customers with goals, patience, trust, and quirks. - [Agentic Eyes reports & evidence](https://docs.qualitymax.io/agentic-eyes-reports.md): Agentic Eyes reports export as HTML or PDF, with captured observations, browser action receipts, findings, and explicit evidence limits. - [AI & Models](https://docs.qualitymax.io/ai-models.md) - [AI Validation](https://docs.qualitymax.io/ai-validation.md): Validate AI agents and AI-generated work — adversarial attack suites, conversation evals with judge criteria, the hallucination gate for generated tests, and Agentic Eyes persona reviews. - [Azure enterprise deployment requirements](https://docs.qualitymax.io/azure-enterprise-deployment.md): Prepare your organization's Azure environment for a QualityMax VM pilot, covering networking, identity, storage, security, recovery and acceptance. - [Code Review & Gates](https://docs.qualitymax.io/code-review-gates.md): How QualityMax reviews every pull request — the four-gate pipeline (Alpha, Gamma, Delta, Beta), Nexus multi-persona review, review learning, and code-health checks. - [Gate Alpha — AI review & SAST](https://docs.qualitymax.io/code-review-gates-alpha.md): Gate Alpha reviews every pull-request diff with AI and runs a static security scan, applying your review preferences and past dismissals. - [Gate Beta — preview deployment tests](https://docs.qualitymax.io/code-review-gates-beta.md): Gate Beta runs your QualityMax Playwright tests against the live preview deployment once your deploy provider reports it is up. - [Gate Delta — generated scripts](https://docs.qualitymax.io/code-review-gates-delta.md): Gate Delta runs QualityMax-generated test scripts matched to your repo's framework on every PR, with soft warnings instead of hard blocks. - [Gate Gamma — native test suite](https://docs.qualitymax.io/code-review-gates-gamma.md): Gate Gamma runs your repository's own Go, Rust, or pytest test suite in an isolated cloud sandbox and hard-blocks the PR on failures. - [Verdicts & merge blocking](https://docs.qualitymax.io/code-review-gates-verdicts.md): How the four pipeline gates combine into one GitHub check conclusion, and which failures actually block a merge. - [Core concepts](https://docs.qualitymax.io/core-concepts.md): The hierarchy, the two usage meters, where tests run, the pull-request gates, and what counts as evidence. - [Deployment](https://docs.qualitymax.io/deployment.md): Understand QualityMax hosted deployment and the Sovereign private-preview model. - [Discovery & Generation](https://docs.qualitymax.io/discovery-generation.md) - [Evaluation modes](https://docs.qualitymax.io/evaluation-modes.md): Agentic Eyes evaluation modes — conversion, pre-signup intent, information discovery, and authenticated task — and why passing one explicitly beats letting it default. - [Evidence & Trust](https://docs.qualitymax.io/evidence-trust.md): Interpret execution artifacts, verified-fix verdicts, and signed exposure receipts. - [Execution](https://docs.qualitymax.io/execution.md) - [Export QualityMax with QTML](https://docs.qualitymax.io/export-qtml.md): Export QualityMax projects and test intent as portable QTML text, validate the artifact, re-import it, and plan a reversible exit. - [Free & open source](https://docs.qualitymax.io/free-and-open-source.md): Choose standalone QualityMax tools and understand their licenses, account requirements, and provider costs. - [Run Playwright tests on every pull request with GitHub Actions](https://docs.qualitymax.io/guides/playwright-tests-github-actions.md): A copy-paste GitHub Actions workflow for Playwright, how to make it a required check, and how to add QualityMax-managed suites and generated tests to the same pull request gate. - [Test a FastAPI app with pytest](https://docs.qualitymax.io/guides/test-fastapi-app.md): Write pytest tests for a FastAPI service with TestClient and dependency overrides, cover async code, run the suite on every pull request, and load-test it from its OpenAPI schema. - [Add end-to-end tests to a Next.js app](https://docs.qualitymax.io/guides/test-nextjs-app.md): Set up Playwright for a Next.js App Router project with a production build, stable locators, signed-in tests, CI, and preview-deployment checks, then extend coverage with QualityMax. - [Run Playwright tests against a Vercel preview deployment](https://docs.qualitymax.io/guides/test-vercel-preview-deployments.md): Run an existing Playwright suite against Vercel preview deployments, manually or on every successful deployment with GitHub Actions, reach protected previews, and understand the current QualityMax webhook setup limitation. - [Hallucination gate](https://docs.qualitymax.io/hallucination-gate.md): The hallucination gate executes every AI-generated test against the real target and classifies it — verified, real bug, hallucinated (quarantined), or inconclusive — from deterministic signals. - [QualityMax documentation](https://docs.qualitymax.io/index.md): Product documentation for QualityMax — the independent system that verifies what AI writes. - [Integrations](https://docs.qualitymax.io/integrations.md): Connect QualityMax to coding agents through Smithery or direct MCP setup, then use Slack, your repository, and CI as reviewable testing surfaces. - [Introduction](https://docs.qualitymax.io/introduction.md): What QualityMax is, the verification loop it runs, and what each product line does. - [Memory & Grounding](https://docs.qualitymax.io/memory-grounding.md): How QualityMax carries durable project memory across agent sessions and keeps AI-generated tests, learnings, and repairs tied to cited, verifiable evidence. - [Example: repairing with vs. without memory](https://docs.qualitymax.io/memory-grounding-example.md): A worked comparison of a test-repair task with and without project memory recall, and what actually drives MCP token and call cost. - [The memory graph](https://docs.qualitymax.io/memory-grounding-graph.md): The navigable graph of agent sessions, tool calls, and verified evidence in the Sessions tab — how it's structured and how it stays authorized and bounded. - [Grounded learnings & evidence](https://docs.qualitymax.io/memory-grounding-grounded-learnings.md): How QualityMax keeps generated tests, project learnings, and discovered facts tied to evidence a scanner or a passing run actually observed. - [Project memory](https://docs.qualitymax.io/memory-grounding-project-memory.md): The get_project_memory MCP tool — how QualityMax recalls verified scripts, prior agent sessions, and provenance for a project before generating or repairing tests. - [Recall & retrieval](https://docs.qualitymax.io/memory-grounding-recall-intents.md): How QualityMax retrieves relevant past learnings with hybrid lexical and vector search, and why recalled memory is always treated as untrusted advisory text. - [Reuse & rediscovery](https://docs.qualitymax.io/memory-grounding-reuse.md): How QualityMax detects when work is being repeated across sessions — duplicate negotiations, proven locators, recurring failures — instead of silently redone. - [Migrate Cypress to Playwright](https://docs.qualitymax.io/migrate-cypress-to-playwright.md): Move Cypress end-to-end tests to Playwright with runnable before-and-after examples, stable locators, fixtures, and parity checks. - [Migrate Gatling to k6](https://docs.qualitymax.io/migrate-gatling-to-k6.md): Convert Gatling Scala simulations to runnable k6 JavaScript in QualityMax and verify requests, users, pacing, and checks. - [Migrate JMeter to k6](https://docs.qualitymax.io/migrate-jmeter-to-k6.md): Convert JMeter JMX plans to reviewable k6 JavaScript in QualityMax, with runnable before-and-after examples and parity checks. - [Migrate Locust to k6](https://docs.qualitymax.io/migrate-locust-to-k6.md): Convert Locust HttpUser tasks to runnable k6 JavaScript in QualityMax and preserve host, task intent, pacing, and assertions. - [Convert Playwright to k6](https://docs.qualitymax.io/migrate-playwright-to-k6.md): Turn Playwright API calls into runnable k6 load tests in QualityMax while keeping browser checks separate from protocol load. - [Migrate Selenium to Playwright](https://docs.qualitymax.io/migrate-selenium-to-playwright.md): Move Selenium WebDriver tests to Playwright with runnable examples, auto-waiting locators, page objects, and parity validation. - [Migrate TestRail to QualityMax](https://docs.qualitymax.io/migrate-testrail.md): Import TestRail cases into QualityMax, map projects and runs, sync execution results, and keep a documented path back out. - [Mobile](https://docs.qualitymax.io/mobile.md): Collect responsive browser evidence, review mobile readiness, and maintain supported mobile flows. - [Nexus persona catalogue](https://docs.qualitymax.io/nexus-personas.md): The eight Nexus crew personas — QA, strategy, pragmatism, security, performance, DevSecOps, UI/UX, and accessibility — with roles, expertise, and deference. - [Nexus multi-persona review](https://docs.qualitymax.io/nexus-review.md): Nexus reviews code as a crew of AI specialists running in parallel — different roles, different blind spots, one aggregated report you can export and share. - [Partner / White-label](https://docs.qualitymax.io/partner-white-label.md): Plan an integration using QualityMax APIs, project workflows, and supported deployment options. - [Performance](https://docs.qualitymax.io/performance.md): Create, review, run, and interpret k6 load tests with QualityMax. - [I use a coding agent](https://docs.qualitymax.io/persona-coding-agent.md): Connect a coding agent to QualityMax and keep the verification loop close to your work. - [I’m a QA engineer](https://docs.qualitymax.io/persona-qa-engineer.md): Organize test cases and connect QTML, TestRail, execution evidence, and coverage. - [I have a repo](https://docs.qualitymax.io/persona-repository-owner.md): Bring a repository into QualityMax, generate a focused test, and add an evidence-based gate. - [I’m evaluating for a team](https://docs.qualitymax.io/persona-team-evaluator.md): Evaluate Sovereign, the security model, compliance needs, and on-premises deployment fit. - [Personal coding seats](https://docs.qualitymax.io/personal-coding-seats.md): Run QualityMax work through your Claude Code or Codex subscription, with clear consent, isolated execution, serialized seat turns, and costs. - [Quickstart: CLI and local agent](https://docs.qualitymax.io/quickstart-cli.md): Install qmax, authenticate in a browser, run behind a firewall, and use qmax-code to inspect a passing run. - [Quickstart: coding agent with MCP](https://docs.qualitymax.io/quickstart-mcp.md): Connect a coding agent, crawl a safe target, generate a test, run it, and inspect the evidence. - [Quickstart: web app](https://docs.qualitymax.io/quickstart-web-app.md): Sign up, import a repository, generate test cases, add a first CI gate, and inspect execution evidence. - [Reference](https://docs.qualitymax.io/reference.md) - [MCP tool reference](https://docs.qualitymax.io/reference/mcp.md): A generated, public-safe reference for QualityMax MCP tools, including parameters, responses, and worked examples. - [Agentic evaluation](https://docs.qualitymax.io/reference/mcp/agentic-evaluation.md): Evaluate generated work with agent and persona reviews. - [Catalog & account](https://docs.qualitymax.io/reference/mcp/catalog.md): Discover tools and inspect the current account context. - [Discovery & generation](https://docs.qualitymax.io/reference/mcp/discovery-generation.md): Crawl applications and produce test coverage. - [Execution](https://docs.qualitymax.io/reference/mcp/execution.md): Run tests and inspect execution results. - [Integrations](https://docs.qualitymax.io/reference/mcp/integrations.md): Connect QualityMax work with supported developer workflows. - [Mobile](https://docs.qualitymax.io/reference/mcp/mobile.md): Audit mobile experiences and work with mobile flows. - [Performance](https://docs.qualitymax.io/reference/mcp/performance.md): Create, migrate, run, and report on k6 performance tests. - [Portability](https://docs.qualitymax.io/reference/mcp/portability.md): Import, validate, convert, and export portable test formats. - [Quality & code review](https://docs.qualitymax.io/reference/mcp/quality-review.md): Review code, coverage, and generated test quality. - [Test management](https://docs.qualitymax.io/reference/mcp/test-management.md): Projects, repositories, test cases, and scripts. - [Review learning & preferences](https://docs.qualitymax.io/review-learning.md): Control what the QualityMax reviewer checks, per repo and globally — and how dismissed findings are remembered so false positives never resurface. - [Security](https://docs.qualitymax.io/security.md) - [Self-Healing](https://docs.qualitymax.io/self-healing.md): Review, validate, and apply QualityMax repair suggestions when an automated test fails. - [Slack](https://docs.qualitymax.io/slack.md): Invite the QualityMax bot to any Slack channel to retrieve authorized test records, switch Nexus specialists, and approve runs, edits, and incident healing. - [Code health & tech-debt checks](https://docs.qualitymax.io/tech-debt-scan.md): The QualityMax Code Health GitHub check posts line-level tech-debt annotations on every PR — complexity, duplication, and maintainability findings as review comments. - [Test Management](https://docs.qualitymax.io/test-management.md): Organize QualityMax projects, test cases, automation scripts, imports, and portable test intent without losing source traceability. ## Sitemap - [HTML sitemap](https://docs.qualitymax.io/sitemap-index.xml)