skills.rest
janewilkin
janewilkinCOMMUNITY
Explore all skills by this author →
GitHub

prompt-review

Analyze and grade AI prompts in application source code.

Updated Feb 17, 2026
Search Tags:#code analysis#prompt engineering#ai quality#llm optimization#prompt review#cost efficiency
One-click install
npx skills add https://github.com/janewilkin/bismuth-digital-llc-web --skill prompt-review-janewilkin
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: prompt-review
Source: https://github.com/janewilkin/bismuth-digital-llc-web/tree/main/.claude/skills/prompt-review
Command: npx skills add https://github.com/janewilkin/bismuth-digital-llc-web --skill prompt-review-janewilkin

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires python -c, git diff *, git log *, wc *, and includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill identifies and improves AI prompts within your codebase, ensuring they are clear, specific, and cost-efficient, while mitigating risks like prompt injection.

Core Features & Use Cases

  • Prompt Discovery: Automatically finds AI prompts across your source code.
  • Quality Grading: Evaluates prompts against a 6-dimension framework (Clarity, Specificity, Context, Structure, Robustness, Efficiency).
  • Actionable Rewrites: Provides concrete suggestions and code diffs for prompt improvement.
  • Use Case: Integrate this into your CI/CD pipeline to automatically flag and suggest fixes for poorly performing or risky AI prompts before they impact your application's quality or cost.

Quick Start

Run the prompt-review skill to analyze all prompts in the current project directory.

Frequently Asked Questions about prompt-review

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

FAQPage Schema
How do I analyze and grade AI prompts embedded in my source code?▼

This skill evaluates AI prompts in source code against a six-dimension framework covering clarity, specificity, context, structure, robustness, and efficiency. It prioritizes improvements based on impact and risk while excluding prompts found in the .claude/ directory.

What is the best way to optimize LLM prompts for cost efficiency and security?▼

Optimizing LLM prompts for cost and security requires evaluating their structure and robustness against prompt injection. This skill analyzes your codebase to identify inefficient prompts and generates actionable code diffs to mitigate risks and lower operational costs.

Can I integrate prompt analysis into my CI/CD pipeline to catch risky AI prompts?▼

Yes, you can integrate this prompt review process into your CI/CD pipeline to automatically flag and suggest fixes for poorly performing or risky AI prompts. This prevents bad prompts from impacting application quality or cost before deployment.

Does the prompt review tool analyze all files in the project directory?▼

The prompt review tool analyzes source code across your project directory but explicitly excludes prompts located within the .claude/ directory. It leverages git diff and git log commands to discover and evaluate prompts within your codebase.

How do I get actionable code diffs for improving poor AI prompts?▼

To get actionable code diffs for improving poor AI prompts, run the skill to analyze your project. It grades each prompt against its quality framework and provides concrete, prioritized rewrites formatted as code diffs based on impact and risk.

Related Skills

Explore More Software Engineering Skills→
borkwebborkweb
7

writing-commits

Generate and format git commit messages using Conventional Commits standards.

Community
Intermediate
viktorbezdekviktorbezdek
10

llm-cost-optimization

Diagnose LLM spending and apply context, prompt, and model optimizations with a quality gate.

Community
Advanced
DevelopersGlobalDevelopersGlobal
65

ci-cd-pipelines

Automate CI/CD pipelines with quality gates and deployment processes.

Official
Advanced
stephanjstephanj
9

prompt-engineer

Design and optimize LLM prompts with evaluation frameworks and few-shot learning.

Community
Advanced