Spec Kit
GitHub (Microsoft)GitHub's toolkit for spec-driven development, which works with coding agents rather than being one. A structured specify, plan, tasks and implement workflow produces Markdown artifacts that feed each phase, giving an agent structured context instead of ad-hoc prompts. Its CLI scaffolds those files and orchestrates more than thirty agents.
Openness
5 high confidence- license
- MIT(OSI)
- source
- public
- core features withheld
- no
Fully MIT and free: the whole Spec-Driven Development toolkit is published, with no paid tier, enterprise directory or license key.
- https://github.com/github/spec-kit recorded 2026-08-13
MIT license, full public source for the Spec-Driven Development toolkit, 127.2k stars and 11.4k forks; no paid tier or enterprise directory
- https://raw.githubusercontent.com/github/spec-kit/main/LICENSE recorded 2026-09-29
The LICENSE file is the MIT License: "MIT License" and "Copyright GitHub, Inc." with the standard grant "Permission is hereby granted, free of charge, to any person obtaining a copy".
- https://raw.githubusercontent.com/github/spec-kit/main/pyproject.toml recorded 2026-09-29
The package manifest bundles the whole toolkit into the free CLI: "Bundle core assets so `specify init` works without network access"; searched it and the README for pricing, license key and premium tier: none.
- https://raw.githubusercontent.com/github/spec-kit/main/README.md recorded 2026-09-29
The README describes the project as public and freely licensed: "Spec Kit is an open source toolkit that gives AI coding agents structured" processes and states "Spec Kit is [MIT licensed](./LICENSE)."
Adoption
2 medium confidenceAdoption is measured by PyPI downloads of specify-cli, whose version matches the repository's latest tag, confirming it is a live distribution channel rather than a stale mirror. A download count is preferred here over the repository's star count.
- https://pypistats.org/api/packages/specify-cli/recent recorded 2026-08-12
76,423 downloads in the trailing 30 days for specify-cli
Capability
4 medium confidenceMethodology/orchestration layer over coding agents rather than a code-writing engine itself.
- https://github.blog/ai-and-ml/generative-ai/spec-driven-development-with-ai-get-started-with-a-new-open-source-toolkit/ recorded 2026-09-29
The GitHub Blog introduces the toolkit as "a new open source toolkit" and walks the phases "Specify", "Plan", "Tasks" and "Implement".
- https://github.com/github/spec-kit recorded 2026-08-13
'Define what to build before building it - with any AI coding agent'; the specify CLI scaffolding and chained Markdown artifacts; no benchmark figure
- https://raw.githubusercontent.com/github/spec-kit/main/docs/reference/integrations.md recorded 2026-09-29
The integrations reference says "The Specify CLI supports a wide range of AI coding agents" and tabulates them, including Claude Code, Codex CLI, Cursor and Gemini CLI.
- https://raw.githubusercontent.com/github/spec-kit/main/README.md recorded 2026-09-29
The README defines the workflow: "Constitution once per project; specify → plan → tasks → implement → converge per feature." and installs the CLI with "uv tool install specify-cli".
Verified 2026-09-29, except some scores, still awaiting re-confirmation