# dot•requirements docs

> Documentation for dot•requirements — readable, testable, AI-native requirements that live alongside your code.

dot•requirements is a requirements tracking system that treats requirements as
discrete, testable data. It gives AI coding assistants a CLI-based requirements
workflow and connects chat assistants (Claude, ChatGPT) through a remote MCP
connector.

## Documentation pages

- [Introduction](https://docs.dotrequirements.io/): One source of truth for what your software should do — human-readable, testable requirements shared by your team and AI agents, with automatic test coverage.
- [Getting Started](https://docs.dotrequirements.io/getting-started)
- [For Developers](https://docs.dotrequirements.io/getting-started/developers): Set up dot•requirements as a developer — requirements as Markdown in your codebase, referenced directly in test descriptions, with automatic coverage and no BDD framework.
- [For Product Managers](https://docs.dotrequirements.io/getting-started/product-managers): Get started as a product manager — write clear, testable requirements with AI help, keep one source of truth, and see which requirements are actually tested.
- [For AI-First Builders](https://docs.dotrequirements.io/getting-started/ai-first-builders): Set up dot•requirements for AI-first building — give your coding agent structured requirements so it writes the right code and tests, starting from one init command.
- [Concepts](https://docs.dotrequirements.io/concepts)
- [How It Works](https://docs.dotrequirements.io/concepts/how-it-works): How dot•requirements works: compose requirements, sync to the cloud, pull them into your codebase as Markdown, reference them in tests, and track coverage automatically.
- [Documents](https://docs.dotrequirements.io/concepts/documents): Documents in dot•requirements combine prose, diagrams, and embedded requirement blocks — familiar product docs with structured, testable requirements inside.
- [Requirements](https://docs.dotrequirements.io/concepts/requirements): What makes a good requirement — capturing user-facing behavior, business rules, and edge cases as discrete, testable statements your tests can reference.
- [Discovery](https://docs.dotrequirements.io/concepts/discovery): Discovery boards give teams a structured place to explore a problem space together before writing requirements — surface unknowns early and build shared understanding.
- [Coverage Tracking](https://docs.dotrequirements.io/concepts/coverage-tracking): Requirement-level test coverage — track which requirements have tests validating them, so your team sees whether features are tested, not just lines of code.
- [Spec-Driven Development](https://docs.dotrequirements.io/concepts/spec-driven-development): Spec-driven development connects requirements, tests, and implementation in every iteration — enough structure to keep AI-accelerated teams oriented, without heavyweight process.
- [Tools](https://docs.dotrequirements.io/tools)
- [CLI](https://docs.dotrequirements.io/tools/cli): The dotreq CLI manages the requirements workflow: initialize projects, sync with the cloud, validate files, report coverage, and configure AI integrations.
- [Codebase to Spec (Alpha)](https://docs.dotrequirements.io/tools/cli/codebase-to-spec): Codebase-to-spec reads an existing codebase and writes a behavioral specification for it — useful before refactoring legacy code or when onboarding to an unfamiliar system.
- [Test Harness](https://docs.dotrequirements.io/tools/test-harness): Track requirement coverage from your tests with any JavaScript runner — Vitest, Jest, Playwright — plus pytest and JUnit support for Python and Java teams.
- [Web](https://docs.dotrequirements.io/tools/web)
- [Attachments](https://docs.dotrequirements.io/tools/web/attachments): Keep the material around a spec — whiteboards, designs, reference pages — attached to the documents and discovery boards it belongs to.
- [Mockups](https://docs.dotrequirements.io/tools/web/mockups): Lofi wireframe mockups — hand-drawn-style sketches of screens, made in conversation and attached to the spec they illustrate.
- [Source Connections](https://docs.dotrequirements.io/tools/web/source-connections): Connect Linear, Notion, Jira, Confluence, GitHub, and Miro so the dot•requirements assistant drafts specs from your existing product material instead of a blank page.
- [AI](https://docs.dotrequirements.io/tools/ai)
- [Agent Workflow](https://docs.dotrequirements.io/tools/ai/agent-workflow): How AI coding assistants — Claude Code, Cursor, Codex, Copilot — drive the dot•requirements workflow through the CLI, guided by context installed with dotreq ai-setup.
- [MCP Server](https://docs.dotrequirements.io/tools/ai/mcp): Connect Claude or another MCP-capable chat app to your dot•requirements workspace — turn conversations into published, testable requirements documents.
- [CI/CD Integration](https://docs.dotrequirements.io/tools/ai/ci-cd): Run the dotreq CLI in GitHub Actions and other CI — authenticate from environment variables for automated, AI-powered requirements workflows.
- [Local MCP (Retired)](https://docs.dotrequirements.io/tools/ai/local-mcp): The local MCP server is retired — coding assistants now use the CLI directly and chat apps connect through the hosted MCP server. How to migrate with dotreq ai-setup.
- [About Us](https://docs.dotrequirements.io/about): Why Popover built dot•requirements — shipping at AI speed created a context problem, and requirements as testable data is how we manage it.

## Related

- [Homepage](https://dotrequirements.io)
- [Discovery blog](https://discovery.dotrequirements.io)
- [npm package](https://www.npmjs.com/package/@popoverai/dotrequirements)