skills.rest
Jawad-Chaudhary
Jawad-ChaudharyCOMMUNITY
Explore all skills by this author →
GitHub

spec-refiner

Read a current spec, apply feedback, and generate an updated spec.

Updated Dec 31, 2025
Search Tags:#workflow#refinement#specification#ai#claude#code-generation#revision
One-click install
npx skills add https://github.com/Jawad-Chaudhary/Hackathone-2-TODO-Spec-Driven-Development --skill spec-refiner-jawad-chaudhary
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: spec-refiner
Source: https://github.com/Jawad-Chaudhary/Hackathone-2-TODO-Spec-Driven-Development/tree/main/Phase02/.claude/skills/spec-refiner
Command: npx skills add https://github.com/Jawad-Chaudhary/Hackathone-2-TODO-Spec-Driven-Development --skill spec-refiner-jawad-chaudhary

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Iterate on specifications to improve reliability of Claude-driven code generation by incorporating feedback and detected issues.

Core Features & Use Cases

  • Read a current spec, apply feedback, and produce an updated spec.
  • Validate improved specs against common error scenarios in AI-generated code.
  • Use Case: When Claude outputs need refinement after failed builds or failing tests.

Quick Start

Provide the initial spec.md and feedback from the previous Claude output to generate a refined spec.

Frequently Asked Questions about spec-refiner

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I refine an AI code generation specification after a failed build?▼

To refine an AI code generation specification after a failed build, you provide the initial spec and the build feedback to generate an updated spec. This process applies detected error reports to improve code generation reliability.

What is the best way to iterate on Claude specifications using error reports?▼

Iterating on Claude specifications involves reading the current spec, applying changes based on error reports, and running a refinement script. This generates an updated spec designed to prevent previous code generation failures.

Can I use feedback from failing tests to update my markdown spec for AI code generation?▼

Yes, you can use feedback from failing tests to update your markdown spec for AI code generation. The refinement workflow reads the current spec, applies the test feedback, and produces an updated specification file.

When do I need to revise specifications for Claude-driven code generation?▼

You need to revise specifications for Claude-driven code generation when outputs fail builds or tests. Revising the specification incorporates error feedback to prevent the AI from repeating previous mistakes in subsequent generations.

Does spec refinement require a specific script to process feedback?▼

Yes, specification refinement requires running a dedicated Python script. You execute this script by providing the path to your current specification file alongside the feedback from the previous AI output.

Why does Claude code generation fail even with an initial specification?▼

Claude code generation can fail with an initial specification due to unhandled edge cases or ambiguous requirements. Refining the specification by incorporating detected error scenarios validates the spec against common AI-generated code errors.

Related Skills

Explore More Software Engineering Skills→
jeremylongshorejeremylongshore
1

bloated-skill

Generate structured analysis reports from input data with traceable outputs.

Community
Advanced
yellow-seedyellow-seed

test-driven-development

Guide Red-Green-Refactor cycles for test-driven development workflows.

Community
Basic
jamesjlundinjamesjlundin
32

pr-review

Analyze pull requests and generate structured review reports.

Community
Intermediate
jellydnjellydn
114

pr-review

Apply GitHub PR review comments as code edits and commits.

Community
Intermediate