scrutinize

Trace code paths end-to-end and flag mismatches between claimed goals and observed behavior.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/mbaho10/claude-code-skills-9arm-workspace --skill scrutinize-mbaho10
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: scrutinize
Source: https://github.com/mbaho10/claude-code-skills-9arm-workspace/tree/main/.claude/skills/scrutinize
Command: npx skills add https://github.com/mbaho10/claude-code-skills-9arm-workspace --skill scrutinize-mbaho10

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Outsider-perspective end-to-end review of a plan, PR, or code change. First questions intent and whether a simpler/more elegant approach would achieve the same goal, then traces the actual code path (not just the diff) to verify the change does what it claims. Output is concise, actionable, and every call carries its rationale. Trigger on /scrutinize and proactively whenever the user asks to review, audit, sanity-check, or get a second opinion on a plan, PR, diff, design doc, or proposed code change.

Core Features & Use Cases

  • Outsider perspective: forget who wrote it and why. Read the artifact cold.
  • Trace end-to-end: follow the real execution path beyond the diff to verify behavior.
  • Actionable findings with rationale: concise recommendations tied to evidence.
  • Proactive review: triggers on /scrutinize or user requests to audit plans, PRs, diffs, or designs.

Quick Start

Provide the artifact (plan, PR, or code change) and trigger /scrutinize to begin an end-to-end outsider review.

Frequently Asked Questions about scrutinize

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

FAQPage Schema
How do I perform an end-to-end code review that traces actual execution paths beyond a diff?

End-to-end code review traces the actual execution path beyond the diff, analyzing all call sites and state mutations to verify the change does what it claims. It flags mismatches between claimed goals and observed behavior with evidence-backed findings.

What is an outsider perspective code review and when do I need it?

An outsider perspective code review evaluates a plan or PR cold, ignoring who wrote it and why. You need it when you require a second opinion to sanity-check whether a simpler or more elegant approach would achieve the same goal.

How do I audit a PR to verify claimed goals match observed behavior?

To audit a PR, identify the intent behind the artifact and trace the real code path end-to-end. This verifies whether the change does what it claims, prioritizing issues by impact and delivering concise remediation steps with clear rationale.

Can I use an automated review to sanity-check a design doc or proposed code change?

Yes, you can trigger an automated review proactively on design docs and proposed code changes. It provides an outsider review by questioning intent and tracing the actual execution path to ensure the design does what it claims.

What is the best way to review a plan for simpler or more elegant alternatives?

The best way to review a plan is to read the artifact cold and question whether a simpler or more elegant approach would achieve the same goal. This outsider review traces the actual execution path to deliver concise, actionable findings.

Why does my PR review miss call site mutations and state changes outside the diff?

Standard PR reviews often miss state mutations because they only check the diff. Tracing the actual code path end-to-end, including all call sites and exits, verifies behavior and flags mismatches between claimed goals and observed reality.