sigil

Classify code changes into entity-level ADDED, REMOVED, MODIFIED, MOVED, RENAMED, FORMATTING ONLY, or BREAKING categories.

Updated Mar 23, 2026
One-click install
npx skills add https://github.com/gauravverma/sigil --skill sigil-gauravverma
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: sigil
Source: https://github.com/gauravverma/sigil/tree/main/skills/sigil
Command: npx skills add https://github.com/gauravverma/sigil --skill sigil-gauravverma

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Surface code changes at the level of entities (functions, classes, methods) rather than raw text diffs, enabling faster, more meaningful reviews and navigation.

Core Features & Use Cases

  • Entity-level diffs: classify changes as ADDED, REMOVED, MODIFIED, MOVED, RENAMED, FORMATTING ONLY, or BREAKING.
  • Code search and symbol navigation: find callers, callees, and references to understand code relationships.
  • AI-friendly outputs: provide structured JSON suitable for automation, PR comments, and CI checks.
  • Cross-file change tracking: identify consistent changes across multiple files and modules.

Quick Start

Run sigil diff HEAD~1 --json to obtain a structured, AI-friendly view of entity-level changes.

Frequently Asked Questions about sigil

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

FAQPage Schema
How do I get entity-level diffs for PR reviews instead of raw line-by-line changes?

Entity-level diffs classify code changes as ADDED, REMOVED, MODIFIED, MOVED, RENAMED, FORMATTING ONLY, or BREAKING to provide faster, more meaningful PR reviews than raw text diffs.

What is structural code intelligence and how does it identify callers and callees?

Structural code intelligence finds callers, callees, and references to map code relationships across files, revealing how entities are defined and interact within a codebase.

Can I output structured JSON from code diffs to automate CI checks and PR comments?

You can output structured JSON from entity-level diffs to support automation, PR comments, and CI checks by running a diff command with the JSON flag to obtain AI-friendly views.

How do I track code changes across multiple files and modules consistently?

Cross-file change tracking identifies consistent modifications across multiple files and modules by mapping entity-level classifications and token-level differences throughout the codebase.

Does this code intelligence approach work for codebase exploration and commits?

This approach applies to PR reviews, commits, and codebase exploration, revealing what changed, where entities are defined, and how they relate across files.

What are the limitations of structural diffs compared to traditional line-by-line diffs?

Structural diffs focus on entity-level changes like functions and classes rather than granular line-by-line text differences, meaning they may not surface purely cosmetic or non-entity formatting shifts.