- Home
- Skills
- Feiskyer
- Claude Code Settings
- Spec Kit Skill
spec-kit-skill_skill
- Shell
1.1k
GitHub Stars
1
Bundled Files
2 months ago
Catalog Refreshed
4 months ago
First Indexed
Readme & install
Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.
Installation
Preview and clipboard use veilstrat where the catalogue uses aiagentskills.
npx veilstrat add skill feiskyer/claude-code-settings --skill spec-kit-skill- SKILL.md20.0 KB
Overview
This skill integrates GitHub Spec-Kit into a 7-phase, constitution-driven workflow for spec-driven feature development. It automates phase detection, spec creation, planning, task generation, analysis, and implementation scaffolding, storing artifacts under .specify/specs/. It helps teams enforce a project constitution and move from problem definition to implementable tasks with repeatable scripts.
How this skill works
The skill inspects the repository for the spec-kit CLI and .specify directory to detect the current project and feature phase. It runs Spec-Kit scripts to create numbered feature folders, generate spec.md, plan.md, tasks.md, and other artifacts, and emits JSON outputs (branch name, file paths) for automation. Phase detection logic checks for constitution, specification, clarification, planning, tasks, analysis, and implementation markers to decide next actions.
When to use it
- Bootstrapping a new project with a constitution and formal spec workflow
- Creating technology-agnostic feature specifications and acceptance criteria
- Converting a high-level idea into a dependency-ordered task list
- Enforcing team-wide principles and traceable decision frameworks
- Automating branch and spec file creation for consistent feature intake
Best practices
- Draft a concise constitution (.specify/memory/constitution.md) before many features
- Keep specs technology-agnostic and outcome-focused with testable acceptance criteria
- Limit clarification rounds to high-impact questions (max ~5 per round)
- Generate small, 1–4 hour tasks with explicit dependencies and requirement refs
- Run phase-detection checks before executing scripts to avoid duplicate work
Example use cases
- Initialize a project and create the first feature branch and spec using specify init and create-new-feature.sh
- Run the Clarify phase to surface up to five targeted questions that unblock architecture decisions
- Produce a detailed implementation plan (plan.md, data-model.md, contracts) from a clarified spec
- Auto-generate dependency-ordered tasks.md to feed into sprint planning and CI pipelines
- Validate that implementation choices align with the constitution during plan reviews
FAQ
You need Python 3.11+, Git, and the uv package manager; install the spec-kit CLI or use the one-time uvx invocation.
How does phase detection determine the current work stage?
It checks for .specify presence, constitution.md, numbered spec folders, and expected files (spec.md, plan.md, tasks.md) and searches for markers like Clarifications or incomplete task checkboxes to infer phase.