analyze

Analyze codebase questions by reading multiple files and citing concrete references.

Updated Apr 24, 2026
One-click install
npx skills add https://github.com/RJJJJ/project2 --skill analyze-rjjjj
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: analyze
Source: https://github.com/RJJJJ/project2/tree/main/.codex/skills/analyze
Command: npx skills add https://github.com/RJJJJ/project2 --skill analyze-rjjjj

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Read-only, evidence-based explanations of complex codebase questions, eliminating guesswork and avoiding edits while exposing sources and confidence levels.

Core Features & Use Cases

  • Ranked, evidence-based explanations drawn from multiple files across modules.
  • Explicit evidence-vs-inference separation.
  • Cite concrete file paths and lines to support conclusions.
  • Non-destructive analysis that avoids proposing code edits or changes.

Quick Start

Ask to analyze the repository in read-only mode and return a grounded, ranked synthesis with evidence and sources.

Frequently Asked Questions about analyze

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

FAQPage Schema
How do I get an evidence-based codebase analysis without modifying files?

To get an evidence-based codebase analysis without modifying files, request a read-only repository review. The system reads multiple files across modules to synthesize explanations and cites concrete file paths and lines while keeping edits out of scope.

What is the best way to find concrete file references for complex code questions?

The best way to find concrete file references for complex code questions is to use a ranked synthesis approach. It evaluates multiple files across modules and presents a grounded explanation that explicitly cites the exact file paths and lines supporting each conclusion.

Can I separate proven evidence from inferred logic when analyzing a repository?

Yes, you can separate proven evidence from inferred logic when analyzing a repository. The analysis emphasizes an explicit evidence-vs-inference boundary, ranking explanations and exposing provenance and confidence levels to distinguish verified facts from inferred conclusions.

How to analyze a read-only repository across multiple modules?

To analyze a read-only repository across multiple modules, apply the analysis to the repository context and request a grounded synthesis. It reads across modules, ranks explanations, and provides a non-destructive review with explicit sources.

Does this code analysis approach propose code edits or refactoring changes?

No, this code analysis approach does not propose code edits or refactoring changes. It is a non-destructive tool designed to provide evidence-based explanations and confidence levels, keeping all modifications and edits explicitly out of scope.