How does it calculate the quality score?

It evaluates specs against universal dimensions like scope and testability, as well as type-specific rubrics for APIs, UI components, backend features, and bug fixes.

What platforms does this skill support?

Spec Auditor supports Jira issues, Linear issues, GitHub issues, local Markdown files, and inline text snippets.

What is 'Quick Check Mode'?

It is a lightweight guardrail designed to run before you start coding. It provides a fast readiness grade (A-F) and identifies top gaps without requiring a full audit.

Does it automatically update my tickets?

In Full Audit Mode, the skill provides a preview of the improved spec and offers to write it back to the source platform (Jira, Linear, or local file) once you approve.

Can it learn my project's specific coding standards?

Yes, it automatically scans for local instruction files like CLAUDE.md or project documentation to align its audit with your specific terminology and conventions.

Spec Auditor & Optimizer

Name: Spec Auditor & Optimizer
Author: codacy

bycodacy

0•

Collaboration Tools

Audits and enhances technical specifications for AI-agent readiness to ensure clear, testable, and unambiguous implementation.

Spec Auditor & Optimizer (spec-improve) acts as an essential bridge between planning and coding by analyzing Jira issues, Linear tickets, and local files for quality and clarity. It identifies critical gaps in acceptance criteria, error handling, and edge cases, providing a weighted quality score and the ability to automatically rewrite specs to be 'AI-agent ready.' By acting as a guardrail before implementation starts, it minimizes technical debt and prevents coding errors caused by vague or incomplete requirements.

Key Features

01Automatic rewriting of specifications to include testable acceptance criteria

020 GitHub stars

03Context-aware analysis using local CLAUDE.md and project documentation

04Parent context inheritance to ensure consistency across Epics and Stories

05Automated quality scoring for Jira, Linear, and GitHub issues

06Dual-mode operation: Full Audit for deep reviews and Quick Check for pre-work guardrails

Use Cases

01Reviewing a Jira story to ensure it has all necessary details before a developer starts work

02Rewriting vague bug reports into clear, actionable technical specifications

03Checking local spec files for missing error handling and edge cases during development

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add codacy/codacy-specs spec-improve

For use in Claude.ai and ChatGPT