How does the Spec skill integrate with GitHub?

It can create new issues, view existing issue comments, post formatted product specifications as comments, and manage status labels like 'needs-design' or 'needs-spec'.

What is the difference between exploration and quick mode?

Exploration mode is for deep dives into vague ideas using multiple parallel research agents, while quick mode is an autopilot-driven process for well-defined issues that require a formal spec.

Does this skill help with technical implementation?

The Spec skill focuses specifically on the 'What' and 'Why' of a product. For the technical 'How' and architectural design, it is designed to hand off to the Architect skill.

Can it generate visual diagrams?

Yes, it automatically generates Mermaid flowcharts or sequence diagrams based on the feature's logic and can export a visual HTML summary of the session.

Spec Product Discovery

Name: Spec Product Discovery
Author: phrazzld

byphrazzld

•

Herramientas de Colaboración

Clarifies what to build and why through interactive product discovery, research agents, and automated specification generation.

The Spec skill acts as a virtual product lead within your terminal, bridging the gap between raw ideas and technical implementation. It operates in two modes: a deep exploration mode for vague concepts—utilizing parallel agents for problem research, codebase analysis, and user impact—and a quick autopilot mode for well-defined tasks. By automating the creation of comprehensive product specs in GitHub issues, including user stories, acceptance criteria, and Mermaid flow diagrams, it ensures every feature is validated and well-defined before a single line of code is written.

Características Principales

01Automatic generation of Mermaid diagrams for user journeys and decision trees

02Automated GitHub issue management including creation, commenting, and labeling

03Visual HTML summaries of discovery sessions for stakeholder review

045 GitHub stars

05Parallel agent research covering problem space, web research, and codebase patterns

06Dual-mode discovery: deep interactive exploration or autopilot quick-spec

Casos de Uso

01Conducting automated competitive research and codebase impact analysis for new features

02Standardizing acceptance criteria and success metrics across development teams

03Refining vague feature requests into actionable product requirements and user stories

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add phrazzld/claude-config spec

For use in Claude.ai and ChatGPT

Download Skill