cocoindex

Index project content with semantic vector and full-text search using LanceDB.

56|12|Updated Apr 26, 2026

Search Tags:#project management #indexing #semantic search #code search #content retrieval

Ready to run

npx skills add https://github.com/Neuron-Mr-White/UniPi --skill cocoindex-neuron-mr-white

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: cocoindex
Source: https://github.com/Neuron-Mr-White/UniPi/tree/main/packages/cocoindex/skills/cocoindex
Command: npx skills add https://github.com/Neuron-Mr-White/UniPi --skill cocoindex-neuron-mr-white

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires cocoindex, lancedb, and includes scripts (resource) and references (resource) components.

What problem does it solve?

CocoIndex addresses the challenge of efficiently searching through large amounts of project content, ensuring quick and accurate retrieval of relevant information, especially when dealing with significant codebases or complex document structures.

Core Features & Use Cases

Semantic Code Search: Provides powerful search capabilities leveraging semantic vector search and full-text search, with a focus on understanding code context and intent.
Incremental Indexing: Allows for efficient updating of indexes with minimal impact on system resources, suitable for large-scale codebases with ongoing changes.
Use Case: When you're working on a complex software project and need to quickly find a piece of code or document, CocoIndex helps streamline this process.

Quick Start

Search your indexed codebase semantically for "how authentication works" using the cocoindex_search command.

Frequently Asked Questions about cocoindex

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

FAQPage Schema

How do I perform semantic code search across a large project codebase?▼

Semantic code search across a large codebase uses semantic vector and full-text search to understand code context and intent, enabling fast and precise retrieval of relevant functions and documents.

What is the best way to search project content for how a specific feature works?▼

Searching project content for specific features is best handled by indexing code and documents semantically, allowing you to query natural language phrases like "how authentication works" and retrieve precise matches.

Does semantic project content search require setting up indexing pipelines?▼

Semantic project content search requires setting up indexing pipelines to process your files and utilizes LanceDB for storage to manage the semantic vectors and full-text indexes efficiently.

Can I efficiently update indexes for a codebase with ongoing changes?▼

You can efficiently update indexes for a changing codebase using incremental indexing, which allows for updating search vectors with minimal impact on system resources.

How do I manage storage for semantic vector search in document-heavy projects?▼

Managing storage for semantic vector search in document-heavy projects is handled by utilizing LanceDB, which stores the indexed code and text-based documents for quick retrieval.

Why use semantic vector search instead of standard full-text search for code retrieval?▼

Semantic vector search improves code retrieval by understanding code context and intent, whereas standard full-text search only matches exact keywords, making semantic search better for complex software projects.