Getting Started with Graphify

Fresh

Graphify is an open-source knowledge graph skill for AI coding assistants. It transforms folders of code, documentation, papers, images, and videos into interactive, queryable knowledge graphs — dramatically reducing the tokens your AI assistant needs to understand complex codebases.

What Graphify Does

flowchart LR
    A[Your Files] --> B[Graphify]
    B --> C[graph.html]
    B --> D[graph.json]
    B --> E[GRAPH_REPORT.md]
    C --> F[Interactive Visualization]
    D --> G[AI Assistant Queries]
    E --> H[Human-Readable Audit]

Prerequisites

• Python 3.10 or higher
• One supported AI coding assistant (Claude Code, Codex, Cursor, etc.)

Quick Install

pip install graphifyy
graphify install

PyPI Package Name

The PyPI package is named graphifyy (double-y). pip install graphify installs an unrelated package. The CLI command remains graphify.

What You Get After Install

After installation, Graphify is registered as a skill in your AI coding assistant. You can then run:

/graphify .                    # Build graph of current directory
/graphify query "your question"  # Query the graph
/graphify explain "ConceptName"  # Explain a node

Output Files

Every graphify run produces a graphify-out/ directory:

File	Purpose
graph.html	Interactive, clickable visualization
graph.json	Persistent graph for future queries
GRAPH_REPORT.md	High-degree nodes, community clusters, surprising connections
cache/	SHA256 change detection — only reprocesses modified files

Next Steps

• Installation SOP — detailed setup for all platforms
• Build Your First Graph — step-by-step walkthrough
• Platform Setup — Claude Code, Cursor, Codex integration