Configuration

Fresh

.graphifyignore

Create a .graphifyignore file at your project root to exclude paths from graph extraction. Uses the same syntax as .gitignore.

# .graphifyignore

# Dependencies
vendor/
node_modules/
.venv/

# Build artifacts
dist/
build/
*.generated.py
*.min.js

# Large data files
*.csv
*.parquet
data/raw/

# Test fixtures (optional — include if you want test coverage in graph)
test/fixtures/

Common Patterns

Pattern	Excludes
node_modules/	Node.js dependencies
vendor/	Go/PHP vendor directories
.venv/	Python virtual environments
dist/	Build output
*.generated.*	Auto-generated files
*.min.js	Minified JavaScript

Exclude What You Don't Own

Exclude vendored code, generated files, and third-party libraries. The graph is most useful when it reflects code you wrote and understand.

.gitignore Recommendations

# Keep the graph committed (team members get instant context)
# graphify-out/

# But exclude the cache (each machine rebuilds it locally)
graphify-out/cache/

# Exclude large optional exports
graphify-out/wiki/
graphify-out/obsidian/

Environment Variables

Graphify respects these environment variables when running extraction:

Variable	Purpose	Default
GRAPHIFY_MODEL	LLM model for prose extraction	Platform default
GRAPHIFY_VISION_MODEL	Vision model for image/diagram processing	Platform default
GRAPHIFY_MAX_FILE_SIZE	Max file size in MB to process	10
GRAPHIFY_TIMEOUT	HTTP timeout for remote content fetches (seconds)	30

Security Settings

Graphify includes strict input validation by default:

• Only http:// and https:// URLs are accepted for remote content
• File size limits prevent processing of very large files
• HTTP timeout limits prevent hanging on slow remote fetches
• Path containment: all file operations stay within the specified directory
• HTML-escaped node labels prevent XSS in the visualization

Local-First Processing

Code files are analyzed via tree-sitter AST locally — no code is sent to any external server during AST extraction. Only prose/doc content is sent to your platform's LLM API using your own credentials.

Supported File Types Reference

Code (Tree-sitter AST)

Language	Extensions
Python	.py
JavaScript	.js, .mjs, .cjs
TypeScript	.ts, .tsx
Go	.go
Rust	.rs
Java	.java
C	.c, .h
C++	.cpp, .hpp, .cc
Ruby	.rb
C#	.cs
Kotlin	.kt
Scala	.scala
PHP	.php
Swift	.swift
Lua	.lua
Zig	.zig
PowerShell	.ps1
Elixir	.ex, .exs
Objective-C	.m, .mm
Julia	.jl
Vue	.vue
Svelte	.svelte
Dart	.dart

Documents (LLM Extraction)

.md, .txt, .rst, .docx, .xlsx, .pdf

Media (Vision / Transcription)

.png, .jpg, .jpeg, .webp, .gif, .mp4, .mov, .webm, .mp3, .wav, YouTube URLs