One-click install
npx skills add https://github.com/xz1220/oh-my-kimi --skill analyze-xz1220
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: analyze
Source: https://github.com/xz1220/oh-my-kimi/tree/main/skills/analyze
Command: npx skills add https://github.com/xz1220/oh-my-kimi --skill analyze-xz1220

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It helps you understand what a codebase most likely does for a given question using cross-file, evidence-based reasoning—without editing or implementing changes.

Core Features & Use Cases

  • Read-only deep analysis: Produces an explanation based on what the repository says, separating evidence from inference.
  • Evidence-first ranking: Sorts competing explanations by confidence, and explicitly marks unknowns when the repo cannot conclude.
  • Cross-file traceability: Targets questions that require following contracts, architecture, and behavior across multiple files.
  • When to use: Use it before proposing modifications, when investigating “why” causes, or when you need a grounded understanding of system behavior.

Quick Start

Ask the skill to analyze the repository for the root cause of your issue, requiring file-cited evidence and a ranked explanation with stated confidence.

Frequently Asked Questions about analyze

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

FAQPage Schema
How do I find the root cause of a bug in my code repository without modifying files?

Read-only repository analysis identifies the root cause of bugs by tracing cross-file contracts and behavior without editing files. It provides evidence-based reasoning with confidence-ranked conclusions to explain why failures occur.

What is evidence-based reasoning for debugging and how does it rank explanations?

Evidence-based reasoning for debugging separates repository artifacts from inference to explain system behavior. It sorts competing explanations by confidence rankings and explicitly marks unknown limitations when the codebase cannot conclude.

How do I trace cross-file dependencies to understand a workflow before implementing changes?

Cross-file tracing maps architecture and behavior across multiple files to explain how a workflow works. It requires file-cited evidence and produces a ranked explanation with stated confidence before you propose any modifications.

Can I use read-only repository analysis to investigate why a specific failure happens?

Read-only repository analysis investigates why specific failures happen by examining codebase artifacts. It applies evidence first, separating verified facts from inference, and explicitly marks unknown limitations when the repo lacks conclusive evidence.

What are the limitations of evidence-based code analysis when the repository lacks sufficient context?

Evidence-based code analysis explicitly states unknown limitations when the repository lacks sufficient context to conclude. It separates strict evidence from inference, ensuring you know exactly what remains unverified before acting on the analysis.