Back to list
GitHub Repository Compiles Leaked System Prompts from Anthropic OpenAI Google xAI Cursor and Kimi
Open SourcePrompt EngineeringOpen SourceGitHub Trending

GitHub Repository Compiles Leaked System Prompts from Anthropic OpenAI Google xAI Cursor and Kimi

A trending GitHub repository titled system_prompts_leaks by author asgeirtj has surfaced, curating and publishing extracted system prompts from leading artificial intelligence providers. The collection encompasses system prompts from major AI ecosystem players, specifically highlighting models from Anthropic such as Claude Fable 5.1, Opus 5, Claude Design, and Claude Code; OpenAI entries including ChatGPT GPT-6-Astra and Codex; Google systems featuring Gemini 3.8 Flash, 3.1 Pro, and Antigravity; xAI with Grok and Grok Bot; as well as developer platform Cursor and Moonshot AI's Kimi. According to the project maintainer, the repository provides regularly updated system instructions directly extracted from these diverse production environments, offering a centralized catalog of internal directives.

GitHub Trending

Key Takeaways

  • Broad Provider Coverage: The trending open-source repository system_prompts_leaks by author asgeirtj gathers extracted system prompts across major industry developers including Anthropic, OpenAI, Google, xAI, Cursor, and Kimi.
  • Extensive Model Index: Prompts featured in the collection span advanced and specialized tools, including Anthropic's Claude Fable 5.1, Opus 5, Claude Design, and Claude Code; OpenAI's ChatGPT GPT-6-Astra and Codex; and Google's Gemini 3.8 Flash, 3.1 Pro, and Antigravity.
  • xAI and Specialized Assistants: Beyond the largest foundational model creators, the repository documents system directives for xAI's Grok and Grok Bot, AI coding environment Cursor, and conversational platform Kimi.
  • Continuous Maintenance: The repository maintainer states that the repository is maintained with regular updates as new extractions and instructions become available across these platforms.

In-Depth Analysis

Multi-Platform System Prompt Extraction

The system_prompts_leaks repository on GitHub has gained visibility by providing a centralized repository of internal system prompts extracted across the artificial intelligence sector. System prompts serve as foundational behavioral guidelines configured by developers before end-user interaction occurs, steering model persona, reasoning constraints, safety protocols, and formatting conventions. By compiling instructions extracted directly from active production systems, the repository offers transparency into how various frontier labs and developer toolmakers configure their conversational agents.

Anthropic's configurations in the repository highlight a range of internal variants and tool-specific iterations, cataloging prompts for Claude Fable 5.1, Opus 5, Claude Design, and Claude Code. The inclusion of coding- and design-focused designations alongside core model tiers demonstrates the distinct prompt setups utilized for specialized workflow tasks versus broad conversational applications.

Frontier Models and Coding Integrations

In addition to Anthropic, the compilation features major deployments from OpenAI and Google. OpenAI is represented by prompts extracted from ChatGPT GPT-6-Astra and Codex, capturing both general conversational agent configuration and developer-centric code assistance directives. Google's entries encompass Gemini 3.8 Flash, Gemini 3.1 Pro, and Antigravity, reflecting prompt strategies tailored for different latency and reasoning profiles within the Gemini family and related initiatives.

Furthermore, the repository covers distinct implementations from xAI, Cursor, and Kimi. The inclusion of xAI's Grok and Grok Bot provides visibility into the steering behind xAI's conversational systems. The cataloging of Cursor documents prompt engineering approaches deployed in real-time AI code editors, while Kimi illustrates the guidance instructions used for high-context conversational deployment. The maintainer emphasizes that this compiled database is subjected to regular updates as underlying prompts evolve.

Industry Impact

Visibility into Production Prompt Engineering

The aggregation of system prompts across Anthropic, OpenAI, Google, xAI, Cursor, and Kimi presents significant implications for artificial intelligence research, prompt engineering practices, and AI security. For engineers and researchers, comparing system instructions side by side reveals how top labs enforce boundaries, tone, tool use, and operational constraints across different model architectures and target audiences.

At the same time, the presence of these instructions in a trending public repository underscores the challenge of keeping production system prompts confidential. Because system prompts reside within the model's operational context window, external extraction through specialized prompting remains a recurrent phenomenon across LLM deployments. The continuous updating of repositories like system_prompts_leaks highlights the persistence of extraction techniques and reinforces the reality that production instructions are readily inspected by the developer community.

Frequently Asked Questions

What models and platforms are included in the repository?

The repository includes 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, and Kimi.

Who created the repository and where is it hosted?

The repository is titled system_prompts_leaks, authored by GitHub user asgeirtj, and hosted on GitHub Trending.

How often is the prompt collection maintained?

According to the original repository documentation, the collection is maintained with regular updates as new prompts are extracted and verified.

Related News

OpenVINO on GitHub Trending: An Open Source Toolkit for AI Inference Optimization and Deployment
Open Source

OpenVINO on GitHub Trending: An Open Source Toolkit for AI Inference Optimization and Deployment

A GitHub Trending entry dated November 20, 2026, describes OpenVINO as an open source toolkit for optimizing and deploying AI inference. The entry credits openvinotoolkit and links to the OpenVINO repository on GitHub. Its description identifies two purposes for the toolkit: optimization and deployment. It does not explain how OpenVINO carries out either task or provide examples of its use. The supplied item also does not name a release, describe a new feature, or report benchmark results. The listing is therefore an introduction to the project and its stated purpose, not a release announcement or an assessment of how well the toolkit performs.

Univer Office SDK Brings Editing, Calculation, Collaboration, and File Processing to AI Office Products
Open Source

Univer Office SDK Brings Editing, Calculation, Collaboration, and File Processing to AI Office Products

A GitHub Trending entry from dream-num presents Univer Office SDK as a way to build an AI office inside a product. The description names four capabilities: editing, calculation, collaboration, and file processing. It positions those capabilities for both people and AI agents and says the SDK is built on an extensible open-source core. The central distinction is that Univer is presented as a toolkit for adding office functionality to a product, rather than as a finished office application. The brief entry does not describe how the capabilities are implemented, which file formats are supported, or how AI agents interact with the SDK.

Claude-BugHunter: A Comprehensive Red-Teaming Skill Bundle for Claude Code
Open Source

Claude-BugHunter: A Comprehensive Red-Teaming Skill Bundle for Claude Code

Claude-BugHunter, a trending GitHub project by elementalsouls, introduces a robust skill bundle tailored for Claude Code, specifically designed for bug hunting and external red-team operations. This toolset streamlines security assessments by integrating 82 distinct skills, 15 slash commands, and an extensive library of 681 disclosed-report patterns spanning 24 core vulnerability classes. Furthermore, the bundle enhances security testing by incorporating enterprise identity and infrastructure attack matrices, providing users with a structured framework for complex security analysis within the Claude Code ecosystem.