asgeirtj/system_prompts_leaks
#3 this week
Extracted system prompts from Anthropic - Claude Fable 5.1, Opus 5, Claude Design, Claude Code. OpenAI - ChatGPT GPT-6-Astra, Codex. Google - Gemini 3.8 Flash, 3.1 Pro, Antigravity. xAI - Grok, Grok Bot, Cursor, Kimi and more! Updated regularly.
65.6K stars
10.8K forks
65.6K GitHub watchers
Updated 9/13/2026
Build with Backblaze B2
SDKs, agent skills, IDE extensions, and reference pipelines from Backblaze Labs. All open source.
Loading star history...
Use Cases & Benefits
- Provides a curated collection of system prompt instructions extracted from various AI chatbots and language models.
- Enables developers and researchers to access and analyze real system prompts from leading AI platforms for improved prompt engineering.
- Use for studying and reverse-engineering system prompts to enhance custom AI agent behavior and responses.
- Use for benchmarking and comparing prompt strategies across different AI models like Claude, ChatGPT, and Gemini.
- Use for contributing to or updating a centralized repository of system prompts to support AI research and development.
About system_prompts_leaks
NEW: 23 Aug 2025 OpenAI/gpt-5-thinking.md
System Prompts Leaks
Collection of system message instructions for various publicly deployed chatbots.
Feel free to do PR's.
Please use discussions tabs for discussions not the Issues tab.
Discord username: asgeirtj
X profile: https://x.com/asgeirtj
Star History
Discover Repositories
Search across tracked repositories by name or description