tanav.aiScanLeaderboardResearchBlogGet Started
Open appTry free scan →

asgeirtj/system_prompts_leaks

hookskill
ai models

A curated repository of system prompts collected from major AI chatbots including ChatGPT, Claude, Gemini, Grok, and others. Contains verbatim instructions and rules that these AI models receive before user interactions, organized by provider and model version with regular updates as new prompts are captured.

  • Reference the underlying instructions and behavioral guidelines for various AI models
  • Study differences in how competing AI providers structure their system prompts
  • Track prompt changes and updates across model versions over time
  • Understand the hidden rules and constraints built into popular AI chatbots
  • Research AI agent design patterns through real-world prompt examples
1
✓ SAFE
Risk score 1/100
✓ CRITICAL/HIGH reflect AI-verified findings (false positives excluded) · MEDIUM/LOW/INFO are unverified scanner output
AI-verified (CRITICAL/HIGH):
Findings by checker · 0 high-signal, 1 mostly false-positive (hidden by default)
17 findingsclick to expand
Last scanned: Jul 12, 2026
More servers
google-gemini/gemini-cli85
An open-source AI agent that brings the power of Gemini directly into your terminal.
106k★
skypilot-org/skypilot85
obra/superpowers85
Foundational skill pack by Jesse Vincent now in anthropics/claude-plugins-official. Includes ffuf web-fuzzing/pentest skill. Partial analysis done — full hook and plugin inspection pending. tier=T2
191k★
Significant-Gravitas/AutoGPT85
AutoGPT is the vision of accessible AI for everyone, to use and to build on. Our mission is to provide the tools, so that you can focus on what matters.
186k★
nanocoai/nanoclaw85
A lightweight alternative to OpenClaw that runs in containers for security. Connects to WhatsApp, Telegram, Slack, Discord, Gmail and other messaging apps,, has memory, scheduled jobs, and runs directly on Anthropic's Agents SDK
30k★
google-gemini/gemini-cli85
An open-source AI agent that brings the power of Gemini directly into your terminal.
106k★
Scan your entire org's MCP deployment
2,500+ repos pre-scored. 22% carry CRITICAL findings.