System Prompts and Models of AI Tools — open source GitHub project

System Prompts and Models of AI Tools is a controversial collection of published system prompts, internal tool notes, and AI tool model references.

What System Prompts and Models of AI Tools is

System Prompts and Models of AI Tools is a large and controversial collection of system prompts, internal tool descriptions, and model/tool references for popular AI products: Claude Code, Cursor, Devin, Replit, Lovable, Perplexity, v0, Windsurf, VSCode Agent, and others.

It should not be treated as ordinary documentation or an SDK. It is closer to a leak/archive-style repository that records how AI-tool instructions and internal surfaces look when they become public. Security, legal, and ethics context is unavoidable.

What is inside and how it is used

Content shape

This example shows the project shape and the usual way it is used.

Language: Markdown

## Tool name

- System prompt excerpt
- Internal tool names
- Model notes
- Date or source context
- Related discussion

The project is interesting for prompt-injection research, red teaming, AI security, and agent-product builders. It shows how important hidden system instructions have become and how quickly the community analyzes exposed AI-tool behavior.

Strengths and limits

The limits are central: provenance must be checked, freshness can disappear quickly, and using other companies’ internal prompts may violate service rules or security expectations. This page is for understanding the phenomenon, not copying systems blindly.