What problem it solves

The hidden system instructions that shape AI product behavior are difficult to inspect and change across models, products, and versions. This repository makes collected prompts available in a structured, product-oriented archive for study and comparison.

Key features

• Broad product coverage — The archive covers prompts from major AI providers and products, including Anthropic, OpenAI, Google, xAI, Meta, Perplexity, Cursor, Mistral, Kimi, DeepSeek, and others.

  Source: README, Anthropic, OpenAI

• Version-specific archives — Prompts are organized by model, product, component, and release, preserving access to older and deprecated variants alongside current entries.

  Source: README

• Detailed prompt artifacts — Entries include individual system prompts as well as related tools, skills, agents, classifiers, safety instructions, and injected reminders where the README lists them.

  Source: README

• Regular updates — The project states that its collection is updated daily and provides a recent additions and changes section for tracking additions.

  Source: README

• Community contributions — People with different results can contribute through pull requests or direct contact.

  Source: README

Who it may suit

This archive may suit AI developers, researchers, prompt engineers, and technical readers who want to inspect the system instructions associated with different models and products.

Topics to discuss

• How should leaked system prompts be documented and categorized as products evolve?

• Which prompt differences are most useful to compare across model versions?

• What contribution and review process would best support an expanding daily archive?

Written by ReadmeX AI from the project's README, source and public GitHub data; it may contain mistakes.

Read the original README https://github.com/asgeirtj/system_prompts_leaks#readme

ReadmeX opened this community for a trending project; it is not run by the project. Maintainers can claim it from the community page.