Spec Kitty is an open-source command-line tool for Spec-driven development with code agents. It turns product requirements into repeatable workflows, keeping specifications, plans and tasks inside the repository itself in a dedicated directory, and uses Git worktrees to isolate the implementation work. Its declared flow follows the steps spec, plan, tasks, next, review, accept and merge, and the framework announces support for multiple agents including Claude, Cursor and Gemini. It is one of the six frameworks assessed in From Prompt to Process: a Process Taxonomy and Comparative Assessment of Frameworks Supporting AI Software Development Agents.
Capabilities
The worktree isolation is what distinguishes Spec Kitty from the other specification-driven frameworks in that assessment: agents implement work packages in a controlled way and the result passes through review and acceptance before the merge. The study reads this as inserting human control points into the flow and bringing it closer to established code-review practice, and uses it as its worked example of why the taxonomy’s dimensions overlap rather than partition — the worktrees isolate implementation, which is execution, while requiring review and acceptance before the merge is validation, so the same feature scores on both.
Adoption & Ecosystem
Under the Six-Dimension Process Taxonomy, Spec Kitty scores 2 on specification, 1 on context, 1 on roles, 2 on execution, 2 on validation and 1 on portability — a total of 9 out of 12, the second-highest of the six frameworks assessed and the only one scoring 2 on execution. Alongside BMAD (Breakthrough Method for Agile AI-Driven Development), it is one of the two frameworks the study identifies as treating roles and validation as central rather than leaving them weak, and one of the two it names as giving partial answers to the tension between agent autonomy and governance by inserting human review points. The scores express the study author’s judgement from official documentation, not an independent empirical measurement.
The caveat the study attaches is maturity: Spec Kitty was included at the threshold of the study’s traction filter, and the study describes its traction as still modest compared to the others, which leaves its maturity and adoption open. Its portability is also qualified — although it works with multiple agents, it depends on its own repository conventions and on git worktrees. The study additionally lists Spec Kitty among the installable community frameworks that illustrate the emerging supply chain of commands, skills, templates and workflows, and the supply-chain risk that comes with it.