code-quality-review-all

Audit evaluations under src/inspect_evals and generate results.json and SUMMARY.md.

617|397|Updated Oct 2, 2024
One-click install
npx skills add https://github.com/UKGovernmentBEIS/inspect_evals --skill code-quality-review-all
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: code-quality-review-all
Source: https://github.com/UKGovernmentBEIS/inspect_evals/tree/main/.claude/skills/code-quality-review-all
Command: npx skills add https://github.com/UKGovernmentBEIS/inspect_evals --skill code-quality-review-all

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill automates the process of auditing all evaluations in a repository against a single code quality standard. It consolidates findings into a structured results.json and a SUMMARY.md for executive oversight, enabling traceable remediation.

Core Features & Use Cases

  • Bulk Evaluation Audit: Scans every evaluation under src/inspect_evals and reports consistency with a chosen standard.
  • Structured Outputs: Produces results.json with per-evaluation issues and a SUMMARY.md that highlights trends and most affected evaluations.
  • Governance & Remediation: Maintains fix_status for ongoing issues and documents remediation guidance in topic README, ensuring reproducibility.

Quick Start

  • Run the audit workflow to scan all evaluations and generate outputs in agent_artefacts/code_quality/code-quality-review-all/.

Frequently Asked Questions about code-quality-review-all

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

FAQPage Schema
How do I audit all repository evaluations against a single code quality standard?

To audit all repository evaluations against a single code quality standard, run the bulk audit workflow to scan every evaluation under src/inspect_evals and generate a unified results.json and SUMMARY.md.

What is the best way to track remediation progress for code quality issues across a repository?

Tracking remediation progress for code quality issues is handled by generating a results.json file that preserves existing fix_status for ongoing issues, alongside a SUMMARY.md that highlights trends for executive oversight.

How do I generate a summary report of code quality metrics for multiple evaluations?

Generating a summary report of code quality metrics requires scanning all evaluations under src/inspect_evals to produce a SUMMARY.md that highlights trends, identifies the most affected evaluations, and documents remediation guidance.

Does the code quality audit workflow require a topic README for reproducibility?

Yes, the code quality audit workflow requires topic README generation to document remediation guidance, ensuring reproducibility and traceable governance across all scanned evaluations in the repository.

Where are the structured outputs stored when running an automated code quality audit?

When running an automated code quality audit, the structured outputs including the unified results.json and SUMMARY.md are stored in the agent_artefacts/code_quality/code-quality-review-all/ directory.