cortex-scan

Create hierarchical Cortex memories from files, websites, repos, and APIs.

Updated Apr 27, 2026

Search Tags:#API integration #knowledge graph #file system #decomposition #web fetch #hierarchical memory

Ready to run

npx skills add https://github.com/emanueleielo/cortex --skill cortex-scan

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: cortex-scan
Source: https://github.com/emanueleielo/cortex/tree/main/.claude/skills/cortex-scan
Command: npx skills add https://github.com/emanueleielo/cortex --skill cortex-scan

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Enables the building of detailed, structured Cortex memories for concepts with clear decomposition into sub-parts, facilitating comprehensive knowledge organization.

Core Features & Use Cases

Hierarchical Memory Creation: Recursively ingests folders, websites, repos, or documents to mirror their structure within Cortex.
Multimedia Support: Utilizes web fetches, file reads, and API calls to gather diverse types of source material.
Use Case: Build a Cortex memory of a large codebase or research topic by systematically breaking down the subject into sub-concepts, aiding in understanding and retrieval.

Quick Start

Trigger the scan to ingest your project directory or URL, specifying the depth of decomposition to build a detailed Cortex memory tree.

Frequently Asked Questions about cortex-scan

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

FAQPage Schema

How do I build a hierarchical knowledge graph from a complex codebase?▼

To build a hierarchical knowledge graph from a complex codebase, you can scan your project directory to recursively decompose its structure into detailed Cortex memories. This mirrors the folder hierarchy and cross-links sub-concepts for comprehensive concept mapping.

Can I ingest websites and API responses into a structured memory tree?▼

Yes, you can ingest websites and API responses into a structured memory tree. The process utilizes web fetches, file reads, and API calls to gather diverse multimedia source material and map it hierarchically into Cortex memories.

What is the best way to decompose large research topics into sub-concepts for retrieval?▼

The best way to decompose large research topics into sub-concepts is to systematically break down the subject using configurable depth. This structured decomposition ensures comprehensive knowledge organization and aids in understanding and retrieval.

Does this hierarchical memory creation tool work with local file systems and remote repositories?▼

Yes, hierarchical memory creation works with local file systems and remote repositories. It recursively ingests folders, websites, repos, or documents to mirror their exact structure within your memory graph.

How do I control the decomposition depth when scanning project directories?▼

You control decomposition depth when scanning project directories by specifying your desired depth level during the scan trigger. This configures how recursively the tool breaks down subjects into sub-concepts before building the memory tree.