speckit-specify_skill

This skill generates or updates a feature specification from a natural language description using templates to deliver a complete, testable spec ready for
  • Shell

11

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 dceoy/speckit-agent-skills --skill speckit-specify

  • SKILL.md12.4 KB

Overview

This skill creates or updates a feature specification from a natural-language feature description. It generates a concise branch short-name, ensures the next feature number is correct across remote/local branches and specs directories, and produces a complete spec and quality checklist ready for planning.

How this skill works

Given a user-provided feature description, the skill extracts key concepts (actors, actions, data, constraints), generates a 2–4 word short-name, checks existing branches and spec directories to compute the next feature number, and runs the project script to create the feature branch and initial spec files. It fills the spec template, produces a specification quality checklist, validates the spec against the checklist, and iterates fixes until the spec passes validation or presents up to three critical clarification questions.

When to use it

  • You need a new feature spec from a plain-language request
  • You want to update an existing spec while preserving numbering conventions
  • You need a standardized, testable spec ready for planning
  • You want automated branch creation with spec and checklist generation
  • You need to surface critical open questions before planning

Best practices

  • Provide a clear, single-paragraph feature description including actors and desired outcome
  • Mention constraints or compliance needs up front to avoid [NEEDS CLARIFICATION] markers
  • Accept suggested reasonable defaults when details are not critical
  • Limit scope per request—one feature per spec to keep numbering and validation simple
  • Review any produced [NEEDS CLARIFICATION] prompts promptly to unblock validation

Example use cases

  • Add a user-facing password reset flow described in plain language
  • Create a spec for integrating a third-party payment provider with scope notes
  • Update an existing feature spec to expand accepted file types for uploads
  • Generate a standardized spec and checklist for a new analytics dashboard
  • Produce a spec that defines measurable success criteria for an experiment

FAQ

I will return an error asking for the feature description or ask up to three targeted clarification questions if details significantly affect scope or security.

How many clarification questions will you ask?

At most three critical [NEEDS CLARIFICATION] questions prioritized by impact; I make reasonable defaults for the rest.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational