# Checksum AI ## Docs - [Agent Sessions](https://checksum.ai/docs/agent-sessions/index.md): When you trigger detection, generation, or healing, Checksum creates an agent session — a running instance of the AI agent working on your tests. Sessions run autonomously but allow you to interact when the agent needs your input. - [API Reference](https://checksum.ai/docs/api-reference/index.md): Checksum exposes a public REST API that lets you trigger test runs, poll for results, manage test health, and initiate auto-healing — all from your own scripts, CI pipelines, or internal tools. - [Auto-Healing](https://checksum.ai/docs/auto-maintenance/auto-healing.md): Auto-healing is Checksum's batch repair system for tests that fail in CI and can't be fixed by auto-recovery alone. When tests need code-level changes, Checksum creates an AI agent session that analyzes the failures, fixes the test code, and opens a PR with the healed tests. - [Auto-Recovery](https://checksum.ai/docs/auto-maintenance/auto-recovery.md): Auto-recovery is Checksum's real-time test maintenance system. When a test step fails during execution, the CLI's AI fallback immediately attempts to recover — finding the right element, adjusting timing, or adapting to UI changes on the fly. - [Auto-Maintenance](https://checksum.ai/docs/auto-maintenance/index.md): Tests break. Apps evolve, selectors change, flows get redesigned. Maintaining a large E2E test suite is traditionally one of the biggest pain points in testing. - [CI Integration](https://checksum.ai/docs/checksum-cli/ci-integration.md): Run your Checksum tests automatically in your CI pipeline. This guide covers setup for GitHub Actions and GitLab CI/CD. - [CLI Reference](https://checksum.ai/docs/checksum-cli/cli-reference.md): The Checksum CLI (checksumai) is installed as part of the @checksum-ai/runtime npm package. All commands are run via npx checksumai. - [Checksum CLI](https://checksum.ai/docs/checksum-cli/index.md): The Checksum CLI (checksumai) is the command-line tool for running your generated Playwright tests. It handles environment setup, test execution, result uploading, and integrates with the Checksum dashboard for reporting. - [Reports and Traces](https://checksum.ai/docs/checksum-cli/reports-and-traces.md): Every test run executed through the Checksum CLI uploads results to the dashboard. You can view detailed reports, debug failures with the trace viewer, and track test health over time. - [Coding Agent Integration](https://checksum.ai/docs/coding-agent-integration/index.md): Checksum integrates directly with AI coding agents like Claude Code, Cursor, and other tools via slash commands. Instead of using the cloud-based web app, developers can run /checksum commands in their IDE to detect, generate, and heal tests locally. - [Detect Tests](https://checksum.ai/docs/detect-tests/index.md): Test detection is the first step in the Checksum workflow. Instead of manually writing test specifications, Checksum's AI agent analyzes your application and automatically identifies the most important user flows to test. - [Feature Health Dashboard](https://checksum.ai/docs/feature-health-dashboard/index.md): Track suite health, triage bugs as grouped entities, collaborate with comments, and get notified when test health changes. - [Deep vs Standard Modes](https://checksum.ai/docs/generate-tests/deep-vs-standard.md): Every Checksum pipeline — detection, generation, and healing — supports two modes: Standard and Deep. The mode determines how thorough the AI agent's process is. - [Generate Tests](https://checksum.ai/docs/generate-tests/index.md): Test generation is the core of Checksum — an AI agent autonomously creates production-ready Playwright E2E tests from your test flow definitions. The agent plans, implements, reviews its own work, and verifies the tests actually pass before delivering them as a pull request. - [Story and Test Format](https://checksum.ai/docs/generate-tests/story-and-test-format.md): Checksum generates two files for each test: a story file (.checksum.md) that describes the test in human-readable format, and a test file (.checksum.spec.ts) that contains the actual Playwright code. - [Environment Configuration](https://checksum.ai/docs/getting-started/environment-configuration.md): Checksum supports multiple testing environments, each with its own environment URL, credentials, and configuration. This allows you to test against staging, production, or any other environment. - [Getting Started](https://checksum.ai/docs/getting-started/index.md): When you create a new project in Checksum, you'll be guided through a setup wizard with three steps. Each step can be completed in any order, and you can skip steps to come back later. - [Integrations](https://checksum.ai/docs/integrations/index.md): Route Checksum events — health reports, bug status changes, auto-heal updates, run completions — to your team's chat tools. Slack, Microsoft Teams, Discord, and Google Chat are supported. - [Slack @checksum Bot](https://checksum.ai/docs/integrations/slack-bot.md): Install the Checksum Slack app and mention @checksum in a channel to start test generation from Slack threads. - [Checksum AI — E2E Testing Platform](https://checksum.ai/docs/overview/index.md): Checksum is a continuous quality platform that runs alongside CI/CD: testing, healing, and validating automatically so teams can ship more code without trading speed for reliability. - [Periodic Runs](https://checksum.ai/docs/periodic-runs/index.md): Periodic runs allow you to schedule recurring test executions on a cron-based frequency. - [Git Integration](https://checksum.ai/docs/repo-and-cli-setup/git-integration.md): Checksum connects to your Git provider to read your codebase and deliver tests as pull requests. You can connect via GitHub (GitHub App) or GitLab (access token). - [Repository and CLI Setup](https://checksum.ai/docs/repo-and-cli-setup/index.md): Before using Checksum, you need to set up your test repository and connect it to Checksum. This guide walks you through creating or configuring a repository for your Playwright tests, installing the Checksum CLI, and verifying that everything works. - [Repo Mirror](https://checksum.ai/docs/repo-mirror/index.md): The repo mirror is the core concept behind how Checksum interacts with your repository. Checksum maintains a synchronized view of your test repository — reading your existing tests and codebase, and writing back by opening pull requests. - [Microsoft Entra ID SSO](https://checksum.ai/docs/sso/microsoft-entra-sso.md): OAuth 2.0 / OpenID Connect setup guide for approving Checksum in your Microsoft Entra ID (Azure AD) tenant. Written for IT and security administrators. ## OpenAPI Specs - [openapi](https://checksum.ai/docs/api-reference/openapi.json)