1
In-The-Wild Jailbreak Prompts on LLMs
Offers the largest known collection of 'in-the-wild' jailbreak prompts, gathered from various public platforms like Reddit and Discord.
AI Tool
system_prompts_leaks Review
system_prompts_leaks is an open-source collection of extracted system prompts from various leading AI chatbots and tools, hosted as a GitHub repository.
shipped Jun 10, 2026aifreemium
1Features extracted system prompts from over 10 leading AI models including Anthropic Claude Fable 5, OpenAI ChatGPT 5.5 Thinking, and Google Gemini 3.5 Flash.
2The associated GitHub repository had accumulated over 134,000 stars and 33,700 forks by March 30, 2026.
3Regularly updated, with recent commits including changes to v0's prompt on March 8, 2026, and Anthropic's Claude Sonnet 4.6 prompt on March 4, 2026.
4Provides insights into complex AI configurations, such as Claude 3.7's 24,000-token system prompt leaked in May 2025.
system_prompts_leaks at a Glance
Similar Tools
Compare Alternatives
Other tools you might consider
View on Stork→
2
LLM Jailbreak Prompt Injection Dataset
A unified safety dataset aggregating over 30 public sources for jailbreaks, prompt injection attacks, and harmful instructions, primarily for training LLM guardrails.
3
My AI System Prompt Library
A comprehensive and regularly updated library of system prompts for various AI systems and autonomous agents, maintained through automated workflows.
4
system-prompts-and-models-of-ai-tools
A massive collection of reverse-engineered system prompts and internal configurations extracted from over 30 popular AI platforms and tools.
Connect
𝕏
X / Twitter@github
⌘
GitHubgithub.com/asgeirtj/system_prompts_leaks
overview
What is system_prompts_leaks?
system_prompts_leaks is a collection of extracted system prompts tool developed by Asgeir Tjelta that enables AI researchers, prompt engineers, and developers to understand and analyze the underlying instructions governing AI model behavior. It provides a centralized repository of system prompts from models such as Anthropic Claude, OpenAI ChatGPT, Google Gemini, and xAI Grok. The project, primarily hosted as a GitHub repository, aims to demystify the underlying instructions that govern AI behavior and responses, fostering transparency and trust in AI systems.
quick facts
Quick Facts
| Attribute | Value |
|---|---|
| Developer | Asgeir Tjelta |
| Business Model | Open Source |
| Pricing | Free |
| Platforms | Web (GitHub) |
| API Available | No |
| Integrations | N/A |
| Last Incident | 2026-06-08T00:00:00Z |
| Status Feed Type | official |
| Status Page URL | https://status.openai.com |
features
Key Features of system_prompts_leaks
The system_prompts_leaks project offers a comprehensive, open-source collection of system prompts, providing a unique resource for understanding the internal configurations of leading AI models. Its features are centered around providing access to and insights from these extracted instructions.
- 1Extracted system prompts from Anthropic models: Claude Fable 5, Opus 4.8, Claude Code, Claude Design.
- 2Extracted system prompts from OpenAI models: ChatGPT 5.5 Thinking, GPT 5.5 Instant, Codex.
- 3Extracted system prompts from Google models: Gemini 3.5 Flash, 3.1 Pro, Antigravity.
- 4Extracted system prompts from xAI models: Grok, Cursor, Copilot, VS Code, Perplexity.
- 5Content is updated regularly, ensuring access to recent prompt versions.
- 6Serves as a centralized repository for comparative analysis of AI model instructions.
- 7Offers insights into behavioral rules, tool logic, artifact systems, and anti-jailbreak mechanisms within AI configurations.
use cases
Who Should Use system_prompts_leaks?
system_prompts_leaks is primarily designed for individuals and organizations involved in AI development, research, and application, seeking to understand and leverage the foundational instructions of large language models. Its open-source nature makes it accessible to a broad technical audience.
- 1**Prompt Engineers:** To learn how leading AI companies craft instructions and apply these techniques to improve their own prompts for better AI responses.
- 2**AI Researchers and Safety Experts:** To study how constraints, safety guidelines, and internal rules are implemented in AI models, and to research mitigation strategies for issues like prompt injection vulnerabilities.
- 3**AI Developers and Competitive Analysts:** To compare system prompts across different AI models and tools, gaining a 'public teardown' view of how competitors instruct their underlying models, aiding in evaluating and building AI-powered developer tools.
- 4**Educators and Students:** To gain practical insights into the internal workings of generative AI systems, fostering a deeper understanding of AI behavior, limitations, and personalities.
pricing
system_prompts_leaks Pricing & Plans
The system_prompts_leaks project is an open-source initiative hosted on GitHub, making it entirely free to access and use. There are no direct pricing plans, subscription costs, or usage fees associated with the collection itself. Users can clone, fork, and utilize the repository content without financial obligation. However, it is important to note that while the prompts are free, the cost implications of using extensive system prompts with commercial AI models can be significant. For example, a Claude Sonnet 4.6 system prompt of 4.5K tokens could incur an overhead of $1832/month for 100,000 calls/month, demonstrating that while the resource is free, its application with paid APIs carries associated operational costs.
- 1Free Tier: Full access to all extracted system prompts and repository content via GitHub.
competitors
system_prompts_leaks vs Competitors
system_prompts_leaks distinguishes itself not as an AI tool, but as a critical open-source collection providing transparency into the proprietary 'hidden layer' of system prompts that define AI behavior. Its competitive positioning stems from its focus on direct extraction and comparative analysis of these confidential instructions.
1
Offers the largest known collection of 'in-the-wild' jailbreak prompts, gathered from various public platforms like Reddit and Discord.
Similar to system_prompts_leaks in providing a collection of prompts for various LLMs, but specifically focuses on jailbreak prompts rather than general system prompts. It functions as a raw dataset, which may require more technical effort to utilize compared to a curated tool.
2
LLM Jailbreak Prompt Injection Dataset↗
A unified safety dataset aggregating over 30 public sources for jailbreaks, prompt injection attacks, and harmful instructions, primarily for training LLM guardrails.
Directly competes by offering a comprehensive collection of adversarial prompts, similar to the 'leaks' aspect of system_prompts_leaks, but is explicitly framed as a safety dataset for defensive AI research and requires accepting conditions for access.
3
My AI System Prompt Library↗
A comprehensive and regularly updated library of system prompts for various AI systems and autonomous agents, maintained through automated workflows.
Directly comparable as it focuses on collecting and updating system prompts, similar to system_prompts_leaks, but emphasizes a broader range of AI applications beyond just LLMs (e.g., autonomous agents) and is distributed as a dataset.
4
A massive collection of reverse-engineered system prompts and internal configurations extracted from over 30 popular AI platforms and tools.
This is a very direct competitor, as it explicitly provides 'extracted system prompts' from a wide array of commercial AI tools, mirroring the core offering of system_prompts_leaks in scope and nature.
❓
Frequently Asked Questions
+What is system_prompts_leaks?
system_prompts_leaks is a collection of extracted system prompts tool developed by Asgeir Tjelta that enables AI researchers, prompt engineers, and developers to understand and analyze the underlying instructions governing AI model behavior. It provides a centralized repository of system prompts from models such as Anthropic Claude, OpenAI ChatGPT, Google Gemini, and xAI Grok.
+Is system_prompts_leaks free?
Yes, system_prompts_leaks is an open-source project hosted on GitHub and is entirely free to access and use. There are no direct pricing plans or subscription costs associated with the collection itself. However, using the extracted prompts with commercial AI model APIs will incur costs from those respective API providers.
+What are the main features of system_prompts_leaks?
Key features include extracted system prompts from Anthropic models (Claude Fable 5, Opus 4.8, Claude Code, Claude Design), OpenAI models (ChatGPT 5.5 Thinking, GPT 5.5 Instant, Codex), Google models (Gemini 3.5 Flash, 3.1 Pro, Antigravity), and xAI models (Grok, Cursor, Copilot, VS Code, Perplexity). The content is regularly updated and serves as a centralized repository for comparative analysis of AI model instructions.
+Who should use system_prompts_leaks?
system_prompts_leaks is beneficial for prompt engineers looking to improve their prompts, AI researchers and safety experts studying AI behavior and vulnerabilities, AI developers and competitive analysts comparing model instructions, and educators or students seeking to understand the internal workings of generative AI systems.
+How does system_prompts_leaks compare to alternatives?
system_prompts_leaks differentiates itself by offering a direct collection of extracted system prompts from commercial AI models, providing transparency into typically confidential instructions. While other projects like 'In-The-Wild Jailbreak Prompts on LLMs' focus on adversarial prompts, and 'My AI System Prompt Library' offers a broader range of system prompts, system_prompts_leaks directly competes with projects like 'system-prompts-and-models-of-ai-tools' by providing a similar scope of reverse-engineered system configurations.
More on Stork
Related AI Tools
Other tools in this category, ranked by community signal
S
Soniox
🤖 AI Tools
Soniox is a multilingual speech AI platform offering real-time speech-to-text, text-to-speech, and translation APIs with high accuracy and low latency.
S
Synthflow
🤖 AI Tools
Synthflow is an enterprise-ready voice AI platform that automates phone calls with human-like agents using no-code tools or APIs.
W
Wrestle AI
🤖 AI Tools
Wrestle AI is an AI-powered wrestling training app that analyzes matches and provides instant feedback to help athletes improve their technique.
Copilot
🤖 AI Tools
Microsoft's AI assistant that provides help with various tasks across devices and is expected to integrate with WebMCP for web interactions.
Omnigent
🤖 AI Tools
An open-source meta-harness that orchestrates multiple AI coding agents for streamlined development workflows.
ToneAdapt
🤖 AI Tools
A tone-matching ecosystem that helps guitarists and bassists recreate famous song sounds using their existing gear by providing adapted settings.
For builders
This page is doing a job for someone else’s tool.
AI agents read it. Buyers find it. Backlinks accrue. Your tool can have one too — live in 24 hours, indexed by Claude, ChatGPT, and Perplexity, queryable via MCP.