Where to Find the Documentation for Graphify: Complete File Reference

The documentation for Graphify is stored entirely within the Graphify-Labs/graphify repository under the v8 tag, with the root README.md serving as the primary entry point and specialized guides located in the docs/ directory, AGENTS.md, and ARCHITECTURE.md.

Graphify is an open-source tool that transforms codebases into interactive knowledge graphs. Whether you are installing the CLI for the first time or integrating the tool with AI assistants like Claude or Codex, the documentation for Graphify provides comprehensive guidance directly in the source repository.

Primary Documentation Entry Points

The repository organizes documentation by use case, making it easy to locate exactly what you need.

Root README (README.md)

The README.md file at the repository root delivers the essential quick-start instructions, installation commands, and badge links. This is the first file you should open when exploring Graphify, as it contains the high-level overview and basic usage patterns.

Architecture Overview (ARCHITECTURE.md)

For developers who need to understand the system’s capabilities before adopting it, ARCHITECTURE.md provides a high-level component diagram, lists supported file types, and presents performance benchmarks. This file explains how Graphify processes codebases and what you can expect in terms of resource usage.

Technical Deep-Dive Guides

How It Works (docs/how-it-works.md)

The docs/how-it-works.md file contains the most detailed technical explanation of Graphify’s internal mechanics. According to the Graphify-Labs/graphify source code, this document describes the three extraction passes used to parse code, the community detection algorithms for grouping related components, confidence tagging for relationship accuracy, and the exact graph format specification.

Docker and MCP SQLite (docs/docker-mcp-sqlite.md)

For containerized deployments, docs/docker-mcp-sqlite.md provides instructions for running Graphify in a Docker container and configuring the built-in MCP (Model Context Protocol) SQLite server. This is essential for teams running Graphify in isolated environments or integrating it with MCP-compatible tools.

AI Assistant Integration

Agent Installation Guides (AGENTS.md)

The AGENTS.md file contains platform-specific installation snippets for various AI assistants. If you are configuring Graphify for Claude, Codex, Kilo, or other supported platforms, this file provides the exact commands and configuration blocks needed for each environment.

Skill Definitions (graphify/skill-*.md)

Individual skill definitions for each supported assistant are stored in the graphify/ directory with the naming pattern skill-*.md (for example, graphify/skill-claw.md and graphify/skill-amp.md). These files define how specific AI agents interact with the Graphify command set.

Localization Support

Graphify provides extensive localization through the docs/translations/ directory. Non-English speakers can find the README rendered in dozens of languages, such as docs/translations/README.zh-CN.md for Simplified Chinese. All translation files follow the same structural pattern as the root README.

Runtime Documentation

After executing a graph build, Graphify generates a human-readable report at graphify-out/GRAPH_REPORT.md. This runtime document summarizes god nodes (highly connected components), identifies surprising connections in your codebase, and suggests follow-up questions for exploration. This file is created dynamically and complements the static repository documentation.

Quick Start Commands

Below are the exact commands documented in the repository for common workflows.

Install and register the CLI:

# Install using uv (recommended)

uv tool install graphifyy

# Alternative: pipx install graphifyy

# Register the skill with your AI assistant

graphify install

Build and query graphs:

# Build a graph for the current project

graphify .

# Query from terminal

graphify query "what connects auth to the database?"
graphify path "FastAPI" "ModelField"
graphify explain "RateLimiter"

Python API usage:

import graphify

# Load a generated graph

graph = graphify.load_graph("graphify-out/graph.json")

# Find shortest path between two nodes

path = graph.shortest_path("FastAPI", "ModelField")
print(path)

Summary

• The documentation for Graphify is housed entirely within the Graphify-Labs/graphify repository under the v8 tag.
• Start with README.md for installation and quick-start guidance.
• Consult docs/how-it-works.md for technical details on the three-pass extraction process.
• Use AGENTS.md and graphify/skill-*.md files for AI assistant integration.
• Refer to docs/docker-mcp-sqlite.md for containerized deployments.
• Access localized versions in docs/translations/ for non-English documentation.
• Review the generated graphify-out/GRAPH_REPORT.md after each build for project-specific insights.

Frequently Asked Questions

Where is the official documentation for Graphify hosted?

The official documentation for Graphify is stored directly in the GitHub repository at Graphify-Labs/graphify under the v8 tag. Unlike projects that use external documentation sites, Graphify maintains all guides, installation instructions, and technical specifications as Markdown files within the source tree, ensuring version consistency.

What is the difference between README.md and docs/how-it-works.md?

The README.md provides a high-level overview, quick-start instructions, and basic installation steps suitable for first-time users. In contrast, docs/how-it-works.md offers an in-depth technical explanation of Graphify’s three extraction passes, community detection logic, and confidence tagging systems, targeting developers who need to understand the underlying algorithms.

How do I find installation instructions for my specific AI assistant?

Platform-specific installation instructions are located in AGENTS.md and the individual graphify/skill-*.md files. These documents contain the exact configuration snippets needed for Claude, Codex, Kilo, and other supported assistants, ensuring proper registration of the /graphify command in your specific environment.

Is there documentation available in languages other than English?

Yes. Graphify maintains translations of the primary documentation in the docs/translations/ directory. Files such as docs/translations/README.zh-CN.md provide the complete README content in Simplified Chinese, with additional language variants following the same file naming pattern for broad international accessibility.

{
  "mcpServers": {
    "instagit": {
      "command": "npx",
      "args": ["-y", "instagit@latest"]
    }
  }
}