maproom-search

Convert code chunks into numerical vectors for semantic codebase search.

Updated Nov 9, 2025
One-click install
npx skills add https://github.com/manifoldlogic/claude-code-plugins --skill maproom-search
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: maproom-search
Source: https://github.com/manifoldlogic/claude-code-plugins/tree/main/plugins/maproom/skills/maproom-search
Command: npx skills add https://github.com/manifoldlogic/claude-code-plugins --skill maproom-search

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires maproom, sqlite3, numpy, pandas, and includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

This Skill solves the problem of navigating and understanding large codebases by providing semantic search capabilities that allow users to find code by concept rather than just by text.

Core Features & Use Cases

  • Semantic Code Search: Find code by concept, not just by text, allowing users to explore unfamiliar codebases and find implementations by idea.
  • First-Time Setup: Simple steps to initialize the database, scan the repository, and verify the setup.
  • Search Types: Choose between full-text search and vector search for different types of queries.
  • Context Command: Explore a chunk's relationships after finding it via search, including callers, callees, tests, and documentation.
  • Multi-Repo Search: Search across multiple repositories for code and specifications.
  • Troubleshooting: Guidance for common issues and error recovery steps.

Quick Start

Run the following command to search for code related to 'authentication' in the repository:

maproom search --repo <repo> --query "authentication" --format agent

Frequently Asked Questions about maproom-search

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

FAQPage Schema
How does semantic code search help navigate unfamiliar codebases?

Semantic code search converts code chunks into numerical vectors for conceptual matching, allowing you to find implementations by idea rather than exact text strings when navigating unfamiliar codebases.

How do I search for code concepts across multiple repositories?

You can search across multiple repositories for code and specifications by running the maproom search command, specifying the target repo, query string, and output format to retrieve conceptual matches.

Do I need Python and specific libraries to set up vector search?

Yes, you need Python along with specific libraries including maproom, sqlite3, numpy, and pandas to initialize the database, scan the repository, generate embeddings, and perform vector search.

What is the difference between full-text search and vector search for codebase navigation?

Full-text search matches exact text strings, while vector search matches code concepts by converting chunks into numerical vectors, allowing you to choose the appropriate search type based on your query intent.

Can I explore a code chunk's relationships after finding it via search?

Yes, you can use the context command to explore a chunk's relationships after finding it via search, including its callers, callees, associated tests, and documentation.

What should I do if my repository scan or database initialization fails?

The Skill provides troubleshooting guidance for common issues and error recovery steps to resolve failures during database initialization, repository scanning, or search execution.