Skip to content
Integrations

Installs into the tools you already use.

One graphify install registers the skill everywhere: 17 AI assistants · 8 model backends · 36 languages of source, docs & data · export to Neo4j, FalkorDB, Obsidian and more. On-device, MIT-licensed.

Claude Code
AI assistants
Cursor
AI assistants
GitHub Copilot
AI assistants
Codex
AI assistants
Gemini CLI
AI assistants
Amp
AI assistants
VS Code
AI assistants
Aider
AI assistants
Devin
AI assistants
Kilo
AI assistants
OpenCode
AI assistants
Google Antigravity
AI assistants
Kiro
AI assistants
Trae
AI assistants
Hermes
AI assistants
Kimi CLI
AI assistants
Pi
AI assistants
Anthropic Claude
LLM backends
OpenAI
LLM backends
Google Gemini
LLM backends
DeepSeek
LLM backends
Ollama (local)
LLM backends
Kimi / Moonshot
LLM backends
AWS Bedrock
LLM backends
Azure OpenAI
LLM backends
Markdown & docs
Data sources
PostgreSQL (live)
Data sources
Terraform / HCL
Data sources
Code — 36 languages
Data sources
PDF
Data sources
Office (docx / xlsx)
Data sources
Images
Data sources
Audio & video (transcribed)
Data sources
SQL schemas
Data sources
Neo4j
Graph export
Obsidian
Graph export
Mermaid call-flow
Graph export
FalkorDB
Graph export
GraphML (Gephi / yEd)
Graph export
SVG
Graph export
…and more assistants
graphify install auto-detects yours

Per-assistant setup guides

How Graphify connects to each of the 17 assistants, what you can ask, and the trust model behind every answer.

Three ways to query the graph

MCP server

stdio or shared HTTP — 8 graph tools for Claude Code, Cursor, or any MCP client.

CLI

graphify query, path, explain, and prs — plain-English answers straight from your terminal.

Python package

uv tool install graphifyy — one MIT-licensed package, entirely on-device.

Start in one command

Try it on your own repo.

Get started
84,812 stars on GitHub. Read the code.

Evaluating for a team? See Graphify Enterprise →