skills.rest
Stahldavid
StahldavidCOMMUNITY
Explore all skills by this author →
GitHub

sensegrep

Locate functions and classes by behavior using AI embeddings and tree-sitter AST parsing.

1|Updated Dec 23, 2025
Search Tags:#semantic#tree-sitter#code-search#code-exploration#dedup#ai-embeddings
One-click install
npx skills add https://github.com/Stahldavid/sensegrep --skill sensegrep
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: sensegrep
Source: https://github.com/Stahldavid/sensegrep/tree/main/plugin/sensegrep-plugin/skills/sensegrep
Command: npx skills add https://github.com/Stahldavid/sensegrep --skill sensegrep

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

sensegrep enables semantic and structural code search, allowing users to locate functions and classes by behavior rather than exact text, leveraging AI embeddings and tree-sitter AST parsing.

Core Features & Use Cases

  • Semantic search across codebases to find functions and classes by behavior
  • Tree-shaking output to collapse irrelevant regions and highlight relevant symbols
  • Deduplication of similar logic to surface unique implementations
  • MCP server support for AI agents and automated workflows
  • Use Case: Explore large codebases, identify refactoring candidates, and understand architecture patterns

Quick Start

Search for the authentication function by behavior in a large TypeScript project.

Frequently Asked Questions about sensegrep

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

FAQPage Schema
How do I search a large codebase by meaning instead of exact text?▼

Semantic code search uses AI embeddings to locate functions and classes by behavior. This allows developers to find implementations based on what code does rather than exact string matches across large codebases.

Can I find refactoring candidates by analyzing AST structure across different languages?▼

Yes, tree-sitter AST parsing identifies architectural patterns and refactoring candidates across multiple languages. It analyzes code structurally to deduplicate similar logic and collapse irrelevant regions.

What is the best way to deduplicate similar logic in a codebase?▼

Deduplicating similar logic requires semantic analysis with AI embeddings to surface unique implementations. This approach scores code behavior to identify and collapse redundant functions during code exploration.

Does this semantic code search tool support MCP server integration?▼

Yes, MCP server support is included for AI agents and automated workflows. This enables semantic search and tree-shaking capabilities to be accessed directly through automated code exploration pipelines.

When do I need tree-shaking output during code exploration?▼

Tree-shaking output is needed when exploring large codebases to collapse irrelevant regions and highlight relevant symbols. It streamlines architectural understanding by filtering out unused code paths from search results.

Related Skills

Explore More Software Engineering Skills→
xberg-ioxberg-io
26

parsing-source

Generate syntax trees from source code using tree-sitter.

Official
Intermediate
ZateZate
17

godot-development

Create and debug Godot scenes and GDScript game logic.

Community
Intermediate
plurigridplurigrid
60

tree-sitter

Parse source code into ASTs and extract symbols across multiple languages.

Official
Advanced
barkleesandersbarkleesanders
5

file-search

Search codebases for text patterns and syntax-aware structural code patterns.

Community
Intermediate