x1xhlol/system-prompts-and-models-of-ai-tools
FULL Augment Code, Claude Code, Cluely, CodeBuddy, Comet, Cursor, Devin AI, Junie, Kiro, Leap.new, Lovable, Manus, NotionAI, Orchids.app, Perplexity, Poke, Qoder, Replit, Same.dev, Trae, Traycer AI, VSCode Agent, Warp.dev, Windsurf, Xcode, Z.ai Code, Dia & v0. (And other Open Sourced) System Prompts, Internal Tools & AI Models
System Prompts and Models of AI Tools by x1xhlol is the most comprehensive third-party archive of internal system prompts from commercial AI coding products including Cursor, Claude Code, Devin AI, Lovable, Perplexity, and 25+ others.
Our Review
System Prompts and Models of AI Tools is the most comprehensive archive of commercial AI coding tool internals on GitHub -- 137,000 stars and 34,300 forks as of May 2026. Maintained by x1xhlol under GPL-3.0, it collects extracted system prompts, internal tool definitions, and model configurations for Cursor, Claude Code, Devin AI, Lovable, Replit, Perplexity, Windsurf, v0, Warp, NotionAI, and 20+ other AI coding products in active use.
What System Prompts and Models of AI Tools archives:
- •30+ commercial AI tool system prompts the actual instruction prompts shipped inside Cursor, Claude Code, Devin AI, Cluely, Comet, Codex, Junie, Kiro, Lovable, Manus, Orchids.app, Perplexity, Poke, Qoder, Replit, Same.dev, Trae, Traycer AI, VSCode Agent, Warp.dev, Windsurf, Xcode, Z.ai Code, Dia, v0, Cody, and more.
- •Internal tool definitions the JSON schemas and function signatures that each agent uses to call its built-in tools.
- •Model configurations which models each product uses by default, with version specifics where extractable.
- •Folder-per-product layout clean directory structure with one folder per product, easy to browse and diff.
- •Active maintenance new versions added as products update their prompts.
- •GPL-3.0 licensed the archive itself is open and forkable for derivative analysis work.
- •Community PRs accepts contributions for new tools or updated prompt versions.
System Prompts ecosystem:
- •prompts.chat by f user-facing prompt library (separate concept, complementary use).
- •Awesome MCP Servers by punkpeye registry of MCP servers that some of the archived tools use.
- •Anthropic / OpenAI / Google official prompt guides the canonical prompt engineering references from model providers.
Getting started:
Visit github.com/x1xhlol/system-prompts-and-models-of-ai-tools and browse by product folder. Each folder contains the raw system prompt as a text or markdown file, plus any extracted tool definitions. Useful for studying how production AI products structure their agent instructions, debugging your own prompts against canonical examples, or building competitive products that need to match incumbent quality.
Limitations:
This is a third-party extraction archive -- the prompts are not authorized releases from the vendors. Some entries may be outdated as products iterate frequently. Legal status of using extracted system prompts in your own products is unclear; treat as research material rather than reusable code. The archive does not include the trained model weights or fine-tuned configurations, only the textual instructions and tool definitions. GPL-3.0 license on the archive means downstream forks must also be GPL-3.0.
Our Verdict
System Prompts and Models of AI Tools in 2026 is the canonical source for studying how commercial AI coding products structure their agent instructions. 137,000 GitHub stars, 34,300 forks, GPL-3.0, maintained by x1xhlol, covering 30+ products including Cursor, Claude Code, Devin AI, and Windsurf.
For Vibe Builders curious about how the products they use actually work, this archive is the closest thing to a behind-the-scenes tour. You can read Cursor's instructions to its model, Devin's task-planning prompt, and Lovable's app-generation flow side by side and see the patterns each company uses.
For Developers building competitive AI products, the archive is gold. Studying how incumbents structure their tool definitions, error-recovery instructions, and agent loops shortens the gap between your prototype and production-quality output. The folder-per-product layout makes diffing approaches easy.
Skip this archive if you only need user-facing prompts to copy into ChatGPT or Claude -- prompts.chat by f is the right resource for that. Skip if you need authorized, vendor-blessed prompts -- this archive is third-party extraction. Skip if your project is closed-source commercial; the GPL-3.0 license on the archive itself makes copying directly into a proprietary product complicated.
Frequently Asked Questions
What is System Prompts and Models of AI Tools?
System Prompts and Models of AI Tools is a third-party archive of internal system prompts and tool definitions from commercial AI products, maintained by x1xhlol on GitHub. As of May 2026 it has 137,000 stars and covers 30+ products including Cursor, Claude Code, Devin AI, Cluely, Lovable, Replit, Perplexity, Windsurf, v0, and Warp. The archive is licensed under GPL-3.0.
Are these prompts authorized by the vendors?
No. This is a third-party extraction archive, not an official release from the vendors. The prompts are extracted from runtime traces, leaks, and reverse engineering by the community. Treat the contents as research material for studying AI product design rather than as copyable code for production use, and consider the legal implications before reuse.
How current are the prompts?
Currentness varies by product. The archive accepts community pull requests as products iterate, so popular tools like Cursor, Claude Code, and Devin AI tend to stay roughly up to date. Less-tracked products may have older versions of their prompts. Always check the file's git history to see when a specific entry was last updated before assuming it reflects current behavior.
Can I use these prompts in my own AI product?
Legally unclear. The archive itself is GPL-3.0, which restricts use in closed-source commercial products. The original system prompts were extracted from third-party products and may carry their own copyright or trade-secret concerns. Treat the archive as research material; consult legal counsel before incorporating extracted text into a competing product.
Who maintains the archive?
The archive is maintained by x1xhlol on GitHub, with active community contributions through pull requests. The maintainer reviews PRs adding new tools or updated versions of existing prompts. There is no editorial standard beyond format consistency and basic verification that the prompt is what the contributor claims it to be from the named product.
What is system-prompts-and-models-of-ai-tools?
System Prompts and Models of AI Tools by x1xhlol is the most comprehensive third-party archive of internal system prompts from commercial AI coding products including Cursor, Claude Code, Devin AI, Lovable, Perplexity, and 25+ others.
How do I install system-prompts-and-models-of-ai-tools?
Visit the GitHub repository at https://github.com/x1xhlol/system-prompts-and-models-of-ai-tools for installation instructions.
What license does system-prompts-and-models-of-ai-tools use?
system-prompts-and-models-of-ai-tools uses the GPL-3.0 license.
What are alternatives to system-prompts-and-models-of-ai-tools?
Explore related tools and alternatives on My AI Guide.
Open source & community-verified
GPL-3.0 licensed: free to use in any project, no strings attached. 137,936 developers have starred this, meaning the community has reviewed and trusted it.
Reviewed by My AI Guide for relevance, quality, and active maintenance before listing.
Topics