kitty:annotate

Generate node summaries, semantic tags, and role descriptions for codebase graph annotations.

1|Updated Mar 28, 2026
One-click install
npx skills add https://github.com/Kakise/cartographing-kitties-plugin --skill kitty-annotate
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: kitty:annotate
Source: https://github.com/Kakise/cartographing-kitties-plugin/tree/main/plugins/kitty/skills/kitty-annotate
Command: npx skills add https://github.com/Kakise/cartographing-kitties-plugin --skill kitty-annotate

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

The annotate skill enriches the codebase graph by attaching human-readable summaries, semantic tags, and role descriptions to nodes, enabling meaningful and precise semantic search across the codebase.

Core Features & Use Cases

  • Enrich graph nodes with concise summaries and domain tags to improve search relevance.
  • Provide role descriptions to clarify node responsibilities and ownership.
  • Boost semantic search results for questions like "what handles authentication" or "where is this auth logic implemented" after indexing or major code changes.
  • Orchestrate end-to-end annotation workflows after index updates, and proactively re-annotate when annotation_status indicates pending work.

Quick Start

Run the annotate orchestrator to fetch a batch of pending nodes and generate annotations for them.

Frequently Asked Questions about kitty:annotate

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

FAQPage Schema
How do I add semantic tags and summaries to codebase nodes for better search?

Annotating codebase nodes involves generating summaries, semantic tags, and role descriptions for graph nodes. This process enriches the codebase graph, enabling precise semantic search across your codebase.

What is the best way to improve semantic search results after major code changes?

To improve semantic search after major code changes, run an annotation workflow. The system checks annotation_status, retrieves pending nodes in batches, generates updated summaries and tags, and submits them to refresh search relevance.

Can I process large codebases in parallel when generating code graph annotations?

Yes, processing large codebases in parallel is supported. The workflow coordinates batch retrieval of pending nodes and submits structured per-node annotations concurrently through the MCP tooling bundle.

When do I need to re-annotate a codebase graph?

Re-annotate the codebase graph during initial indexing, after substantial code changes, or proactively whenever annotation_status indicates pending work, ensuring semantic search queries return accurate and relevant results.

How does annotating codebase nodes help find domain logic implementations?

By attaching role descriptions and domain tags to graph nodes, annotation boosts semantic search relevance. This enables precise answers to queries like finding authentication handlers or specific domain logic implementations.

Do I need to index my codebase before running the annotation workflow?

Yes, indexing is required. The workflow calls index_codebase first, then checks annotation_status to retrieve pending nodes, preparing context and submitting structured annotations to enrich the codebase graph.