When should I use the refine-requirements skill?

Use this skill whenever a developer or tester identifies a missing business rule, constraint, or edge case during the RED, GREEN, or REFACTOR phases of TDD.

What documentation standards does this skill follow?

It uses the AI SDLC v3.0 methodology, categorizing requirements into Business Rules (BR), Constraints (C), and Formulas (F) with mandatory discovery metadata.

Does this skill update my source code?

Yes, it helps update both the requirements documentation and the corresponding implementation and test files to ensure they remain in sync.

What is the requirements refinement loop?

It is a process where discoveries made during coding (like missing edge cases) are fed back into the requirements documentation to ensure specs match the implementation reality.

AI SDLC Requirements Refinement

Name: AI SDLC Requirements Refinement
Author: foolishimp

byfoolishimp

学習とドキュメンテーション

Automates the continuous refinement of software requirements by capturing edge cases and business rules discovered during the TDD/BDD implementation process.

概要

This skill implements the requirements refinement loop of AI SDLC v3.0, closing the gap between high-level specifications and technical implementation reality. It allows developers to systematically capture discoveries made during TDD phases—such as race conditions, missing constraints, or complex business logic—and feed them back into the primary documentation. By maintaining living requirements, the skill ensures that technical insights are formalized as updated business rules (BR), constraints (C), or formulas (F), preventing re-discovery and improving long-term project maintainability through rigorous traceability.

主な機能

Generates structured commit messages for refinement history
0 GitHub stars
Standardized capture of edge cases and race conditions
Automated updates to requirements documents with discovery metadata
Continuous feedback loop between implementation and requirements
Maintains requirements traceability logs (YAML)

ユースケース

Capturing new calculation formulas discovered while writing BDD test cases
Updating business rules after discovering a race condition during the GREEN phase of TDD
Formalizing missing performance constraints identified during the REFACTOR phase

概要

主な機能

Generates structured commit messages for refinement history
0 GitHub stars
Standardized capture of edge cases and race conditions
Automated updates to requirements documents with discovery metadata
Continuous feedback loop between implementation and requirements
Maintains requirements traceability logs (YAML)

ユースケース

Capturing new calculation formulas discovered while writing BDD test cases
Updating business rules after discovering a race condition during the GREEN phase of TDD
Formalizing missing performance constraints identified during the REFACTOR phase