TrueFlow - Deterministic Code Visualizer
Unblackbox your code with deterministic truth. Zero-code Python runtime instrumentation with 3D animated execution videos.
Features
- Zero-Code Instrumentation: Automatically trace Python code execution without modifying source files
- Real-Time Trace Viewer: Watch function calls and execution flow in real-time via WebSocket streaming
- 3D Manim Visualizations: Generate beautiful animated videos of code execution using Manim
- Performance Analysis: Identify bottlenecks with flamegraph-style performance views
- Dead Code Detection: Find unreachable or unused code paths
- AI-Powered Explanations: Local LLM integration for explaining code behavior (supports vision models)
- Multiple Export Formats: PlantUML, Mermaid, D2, JSON, and Markdown exports
Commands
| Command |
Description |
TrueFlow: Auto-Integrate into Project |
Add tracing to your Python project |
TrueFlow: Show Trace Viewer |
Open the trace visualization panel |
TrueFlow: Generate Manim Video |
Create an animated execution video |
TrueFlow: Export Diagram |
Export trace as PlantUML/Mermaid diagram |
TrueFlow: Connect to Trace Server |
Start real-time trace streaming |
TrueFlow: Download AI Model |
Download local LLM for explanations |
TrueFlow: Start AI Server |
Start the local llama.cpp server |
TrueFlow: Open AI Chat |
Open AI chat panel for code explanations |
Configuration
| Setting |
Default |
Description |
trueflow.traceDirectory |
./traces |
Directory to store trace files |
trueflow.autoRefresh |
true |
Auto-refresh on new traces |
trueflow.pythonPath |
python |
Path to Python interpreter |
trueflow.socketHost |
localhost |
Host for trace streaming |
trueflow.socketPort |
5678 |
Port for trace streaming |
trueflow.manimQuality |
medium_quality |
Video quality setting |
Quick Start
- Open a Python project in VS Code
- Run
TrueFlow: Auto-Integrate into Project from the command palette
- Execute your Python code normally
- View traces in the TrueFlow sidebar panel
- Generate Manim videos or export diagrams
Requirements
- Python 3.8+
- Manim (optional, for video generation)
- llama.cpp (optional, for AI explanations)
AI Explanation Panel
The AI panel supports:
- Context injection from dead code, performance, and call trace panels
- Vision models (Qwen3-VL) for analyzing screenshots and diagrams
- Conversation history with context awareness
- Local inference via llama.cpp server
Installation
From VS Code Marketplace / Open VSX
Search for "TrueFlow" in the Extensions view.
From VSIX
code --install-extension trueflow-0.1.0.vsix
From Source
cd vscode-extension
npm install
npm run compile
Then press F5 to launch Extension Development Host.
License
MIT
Links
| |
