ghostprint

What This Skill Does

Ghostprint is a sophisticated privacy-focused utility designed to obfuscate your LLM usage patterns by injecting synthetic, behaviorally realistic traffic into your existing AI provider streams. By simulating human-like interaction patterns—such as varied inter-arrival times, multi-turn conversations, and diverse topic shifts—it effectively masks your unique behavioral fingerprint. This prevents third-party model providers from profiling your specific intent, professional domain expertise, or individual cadence based on your interaction history. It leverages 6 distinct personas and over 300 topics across 12 domains, ensuring that the background noise appears consistent with a genuine, albeit eclectic, human user.

Installation

For OpenClaw users, installation is streamlined via the native plugin manager. Run the following command in your terminal: clawhub install openclaw/skills/skills/alarawms/ghostprint Alternatively, you can clone the repository directly into your extensions folder: git clone https://github.com/alarawms/ghostprint ~/.openclaw/extensions/ghostprint Followed by: openclaw plugins enable ghostprint openclaw gateway restart

For standalone usage, simply copy the example configuration, input your API credentials, and run python3 ghostprint.py --install-cron to set up persistent, scheduled execution.

Use Cases

Ghostprint is ideal for power users who want to mitigate data harvesting and profiling by AI model providers. It is particularly useful for researchers, developers, and privacy-conscious professionals who utilize multiple LLMs daily and wish to prevent their personal research habits or coding styles from being analyzed for targeted model training or commercial profiling. It is also an effective tool for maintaining baseline activity on newer, less-used accounts to ensure they remain active and monitored.

Example Prompts

• "@ghostprint fire a noise round now to keep my profile active while I take a lunch break."
• "@ghostprint stats, can you show me how many noise sessions were fired over the last week?"
• "@ghostprint help, what are the current active personas currently enabled for my fingerprint injection?"

Tips & Limitations

To maximize the effectiveness of Ghostprint, ensure your API keys are shared across multiple services if possible, as diversifying your noise across Anthropic and OpenAI-compatible providers yields the best masking results. The Poisson-distributed timing is designed to prevent easy detection by frequency analysis, so avoid manually firing sessions too close together, as this may create "burst" patterns that stand out. Note that while this tool significantly degrades the accuracy of external profiling, it does not hide the raw content of your intentional, user-driven requests. Always remain mindful of sharing sensitive information in your primary, non-noisy requests.