What Spectrace does
A three-minute overview of Spectrace — requirements, acceptance criteria, tests, and PR verification with traceability.
Spectrace is a requirements-management platform for software teams. You capture requirements in one place, and Spectrace uses AI to help extract them from source material, draft acceptance criteria, and generate test cases from those criteria.
When a pull request opens on a repository you have mapped to a project, Spectrace verifies the diff against the requirement the pull request references — in its branch name, title, or description — or, failing that, the requirement its changed files are already linked to. The result is posted back to GitHub as a comment and a check run, so reviewers see how the change lines up with the spec before they read the diff.
Every artefact keeps its provenance: which requirement an acceptance criterion belongs to, which requirement a test case was generated from, and which requirement a pull request was verified against. That chain is what Spectrace calls traceability.
Related
- Quick start — from a new account to a verified pull request.
- Core concepts — the terms the rest of the documentation uses.
- Writing a good requirement — how to write a requirement and its acceptance criteria.
- How PR verification works — how a pull request is linked to a requirement and checked.
- Connect an agent — connect an AI coding agent to Spectrace over MCP.