inspect

Identifies changed code entities and ranks them by structural risk from git diffs and commits.

190|8|Updated Feb 12, 2026
One-click install
npx skills add https://github.com/Ataraxy-Labs/inspect --skill inspect-ataraxy-labs
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: inspect
Source: https://github.com/Ataraxy-Labs/inspect/tree/main
Command: npx skills add https://github.com/Ataraxy-Labs/inspect --skill inspect-ataraxy-labs

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

inspect eliminates the overwhelm of reviewing large Git diffs by automatically identifying which changed entities are most risky, then grouping related changes so reviewers can focus on what matters first.

Core Features & Use Cases

  • Entity-level change classification: classifies each changed entity (text-only, syntax, functional) from your diff.
  • Risk scoring with blast radius: ranks entities by structural risk using cross-file dependency/impact signals, so critical API or behavior changes rise to the top.
  • Commit untangling & grouping: separates independent logical changes within tangled commits into reviewable groups.
  • Optional LLM-powered review: triages entities first to reduce the amount of code sent to the LLM, then reviews the highest-risk entities for real issues.

Quick Start

Use the inspect tool to triage the last commit range (or PR) and return ranked high/critical entities with dependency-based blast radius so you can review the highest-risk changes first.

Frequently Asked Questions about inspect

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

FAQPage Schema
How do I prioritize risky code changes in a large Git pull request?

Triage structural review risk by extracting changed entities from Git diffs, computing blast radius from repository dependency signals, and ranking behavior-changing or public-API-impacting edits over cosmetic noise. This automated scoring surfaces the highest-risk changes first so reviewers focus where it matters.

What is the best way to untangle multiple logical changes within a single Git commit?

Untangle Git commits by separating independent logical changes into distinct reviewable groups. The process extracts changed code entities from the commit range, classifies their change type, and groups independent edits so you can review each logical change in isolation rather than as a mixed diff.

How does blast radius scoring work for code review risk triage?

Blast radius scoring works by computing dependent counts from an entity graph built from repository dependency and impact signals. It evaluates each changed entity's cross-file structural impact, ranking edits that affect critical APIs or core behavior higher than text-only or syntax-only modifications.

Can I use LLM-guided review to analyze high-risk entities after Git diff triage?

Yes, you can perform optional LLM-powered review after triage. The process first triages changed entities to reduce the amount of code sent to the LLM, then passes only the highest-risk entities based on dependency signals for LLM-guided analysis to identify real issues efficiently.

Does Git diff triage work with uncommitted file changes and MCP server formats?

Git diff triage targets PRs, commit ranges, and uncommitted file changes, producing ranked output in CLI, JSON, and MCP-friendly formats. It extracts entities from these diffs and computes structural review risk to support integration across various code review environments.

What are the limitations of using structural risk scoring for code review?

Structural risk scoring relies on repository dependency and impact signals to rank changes, meaning it prioritizes behavior-changing and public-API-impacting edits but may not catch logic errors in isolated, low-dependency code where structural blast radius is minimal despite potential functional issues.