Our take
Activeregular commits and releases- Activity
- 90
- Community
- 78
- Issues
- 83
- Pull requests
- 82
+37 stars in the last 30 days
- Releases more often than 93% of the projects we track
- Merges more pull requests than 92% of the projects we track
- More contributors than 76% of the projects we track
Spec Kitty is an actively maintained, governance-focused workflow tool for teams willing to manage agent work through repository-native missions and explicit review gates.
Good fit if
- Your team uses Git and wants specs, plans, and work status to remain part of the repository.
- You coordinate multiple AI coding agents and need isolated worktrees and defined work-package boundaries.
- You want a human-reviewed workflow, and are comfortable evaluating an active project whose current 4.x line is still a prerelease.
Look elsewhere if
- You need a lightweight tool for one-off edits or small scripts, which the project itself identifies as poor-fit cases.
- You do not use Git or do not want a structured sequence of specification, planning, task, and review steps.
- You need hosted tracker or sync features as part of the open-core project, since those flows were removed from it.
All health signals
| Last commit | 2026-10-20 (today) |
|---|---|
| Commits, last 90 days | 100+ |
| Releases, last 12 months | 100+ (latest v4.0.0rc7, 2026-10-15, pre-release) |
| Contributors | 91 (top contributor: 41% of commits) |
| Issues closed, last 90 days | 38+ (typically closed in 1 day) |
| Pull requests merged, last 90 days | 89 (typically merged in 0 days) |
| Project age | 13 months |
Checked on 2026-10-21 with the GitHub API.
Project history
Daily GitHub snapshots since Oct 7, 2026. History covers stars and forks; health signals are not tracked historically.
Stars
1,724Forks
198Change since Nov 16, 2026: +1 stars·+1 forks
Overview
Spec Kitty is a Python CLI for spec-driven development with AI coding agents, built around Git-native mission artifacts and isolated Git worktrees. It turns product intent into specs, plans, and work packages, then guides implementation through a visible next, review, accept, and merge workflow.
It is aimed at teams that want requirements and governance to stay with the repository rather than disappear into agent sessions. It supports workflows with Claude Code, Codex, Cursor, Gemini, GitHub Copilot, Windsurf, and other coding agents. The default approach keeps people involved in defining intent and reviewing work rather than treating software creation as a black box.
Key Features
- Guides work from product intent through specification, planning, task creation, review, acceptance, and merge.
- Stores mission specifications, plans, work packages, and acceptance criteria in the repository.
- Organizes work packages into lifecycle lanes such as planned, in progress, for review, approved, and done.
- Uses isolated Git worktrees to support parallel agent or developer work.
- Provides a
nextcommand to identify the next action for an agent on a mission. - Includes governance and review gates, with mission retrospectives for learning from completed work.
- Offers integrations such as slash commands or skills for several coding agents and harnesses.
Use Cases
- A development team wants to turn product requirements or issues into trackable work packages with explicit acceptance criteria.
- A team running several coding agents in parallel needs worktree isolation and clear ownership boundaries.
- Engineers want to preserve architecture decisions and review state in Git across separate AI coding sessions.
- A team moving beyond ad hoc prompting needs a repeatable workflow that still leaves review and merge decisions visible to people.
What you need
Detected in the repository
- Node.js (from package.json)
- Python >=3.11 (from pyproject.toml)
- A test suite and automated checks on GitHub Actions
License in plain words
MITpermissive
- Commercial use: yes
- Modify and redistribute: yes
- You must keep: the copyright and license notice
- Share your changes: no
A summary, not legal advice: the LICENSE file is what applies.
Getting Started
Install the CLI and initialize a project:
pipx install spec-kitty-cli
spec-kitty init my-project --ai claude
See the README for setup verification, supported agents, and the full mission walkthrough.
Alternatives
- gsd-pi: GSD Pi plans and implements milestones locally, while spec-kitty organizes spec-driven work packages, reviews, and merge decisions in Git.
- uta: uta routes decomposed tasks to installed agent CLIs and records runs, while spec-kitty centers on Git-backed specifications and isolated worktrees.
- ralph-orchestrator: Ralph Orchestrator structures agent runs as role-based loops with quality gates, rather than spec-kitty's Git-based work-package and review workflow.
| Project | Language | License | Stars | Status |
|---|---|---|---|---|
| spec-kitty | Python | MIT | 1.7k | Active |
| gsd-pi | TypeScript | MIT | 1.3k | New |
| uta | Go | MIT | 0 | New |
| ralph-orchestrator | Rust | MIT | 3.2k | Active |
Considerations
The project is active, with regular commits and releases, and a broad contributor base that is not concentrated in one person. Its latest release is a prerelease, so teams adopting the 4.x release-candidate line should account for qualification work before relying on it as a stable release. The workflow also brings structure and review overhead that may not be worthwhile for small or one-off tasks. Hosted authentication, sync, and tracker flows are no longer part of the open-core project; its core workflow is local-first and repository-based.
Found this useful?
Share it with someone who would like spec-kitty.