Data & Security

Data Flow: Your Data Never Touches OpenClacky's Servers

The OpenClacky Agent runs locally on your machine. When you use any AI model, the communication path is:

Your Machine (OpenClacky Agent) → AI Provider (Anthropic / OpenAI / Kimi / etc.)

Your conversations, files, and code do not pass through the OpenClacky platform. OpenClacky's platform only handles Skill distribution and license verification — it does not proxy, forward, or store any of your conversational data.

If you use a privately deployed AI provider, your data never leaves your internal network.

AI Keys: Direct Passthrough, Zero Logging

The OpenClacky platform offers managed AI Keys (API Keys). These keys connect directly to models hosted on first-tier cloud providers such as AWS and Google. All requests are directly forwarded — OpenClacky does not log or retain any data.

You can verify this by reviewing the full source code in the openclacky GitHub repository.

Telemetry

To understand client activity and usage patterns, the OpenClacky client uploads the following anonymous statistics after each task:

Token usage (input / output tokens)
Model used
Task duration
Client version

This telemetry does not include any task content, conversation text, file paths, or personally identifiable information.

The upload process is fully open-source. See lib/clacky/telemetry.rb.

To disable telemetry, set the environment variable:

export CLACKY_TELEMETRY=0

Enterprise: 100% Private Deployment

For organizations with strict data compliance requirements, we offer a fully private solution:

Zero telemetry, zero outbound data
All data stays within your corporate network
Compatible with self-hosted AI providers