Privacy Policy

Overview

Toonify MCP is open-source software that runs locally on your machine. This page explains what Toonify itself does locally, what the website does, and where third-party services may still be involved in your overall workflow.

Data Processing

Local processing by Toonify: Toonify's own compression and supporting local features run on your machine:

TOON format conversion for supported content
Local caching and prompt-cache coordination
Local statistics tracking

No Toonify-hosted service: Toonify does not operate its own remote API, telemetry pipeline, or hosted dashboard for this project. It does not separately upload your content or local statistics to PCIRCLE AI-operated servers.

Your client and provider still matter: If you use Toonify with Claude Code or another MCP client, that client may still send prompts, files, tool output, or optimized output to the model provider or platform you use. Those network transfers are governed by that client and provider's terms and privacy practices, not by Toonify alone.

Example: When you use Toonify with Claude Code, optimized output may still be passed to Anthropic through your existing Claude Code usage. See Anthropic's Privacy Policy, Claude Code data usage, and applicable terms for your Claude plan.

Website (toonify.pcircle.ai)

No Analytics: Our documentation website does not use tracking, analytics, or cookies.

Google Fonts: The website uses Google Fonts for typography. Google may collect your IP address as part of font delivery. See Google Privacy Policy.

Open Source Transparency

Toonify is MIT-licensed open source. You can:

Inspect all source code at GitHub
Inspect what Toonify does locally and what external dependencies it uses
Modify the code to suit your needs
Use it freely in commercial or non-commercial projects

Your Control

Toonify's compression, caching, and local stats remain on your machine
No separate Toonify telemetry or usage tracking is enabled on this project website
You control whether to use plugin mode, MCP mode, or neither
You control which model client or provider receives downstream prompts and output
Complete source code transparency - view on GitHub

Contact