How hussainasghar/system-prompts-and-models-of-ai-tools works
Project type and purpose
• This repo is a static content collection of AI “system prompts” and accompanying tool-config files, organized by agent or service (v0, Cursor, Manus, Same.dev, Lovable, Devin, Replit, Windsurf, VSCode, plus various open-source prompts).
• Its purpose is to host & share “full official” and open-source prompts, tool definitions (.json), and model metadata (.md), along with a README that serves as the project landing page.
File structure
• Top-level folders are your primary modules:
• v0 Prompts and Tools (Model.md, Prompt.txt)
• Cursor Prompts (.txt)
• Manus Agent Tools & Prompt (Prompt.txt, Agent loop.txt, Modules.txt, tools.json)
• Lovable, Devin AI, Same.dev, Replit, Windsurf, VSCode Agent (each with .txt prompts and some .json tools files)
• Open Source prompts (subfolders: Bolt, Cline, Codex CLI, RooCode – each with Prompt.txt)
• Root contains README.md and LICENSE.md. There are no src/ or runtime code; everything is authored content.
README for additional insights
• Describes the collection (6,500+ lines of prompts & tools), highlights “official” vs “open-source,” and links to external services (ZeroLeaks for security, system-prompts.featurebase.app for feedback, star-history.com for metrics).
Main components and their relationships
a. Content Repository (GitHub)
• Modules (folders) containing prompt files and tool definitions
• README and license at root
b. Build / Publication Pipeline (optional but recommended)
• Static Site Generator (e.g., Docusaurus, Hugo) consumes the repo
• Outputs a browsable site or API for prompt lookup
c. Hosting / Frontend
• GitHub Pages / Vercel or similar, serving the generated site
d. End-Users
• AI engineers, startups, community members consuming prompts & tools
e. External Services
• star-history.com (visualizes repo stars over time)
• ZeroLeaks (AI security audit link)
• Roadmap & Feedback (system-prompts.featurebase.app)
Architectural patterns and design principles
• Modular content organization: each agent/type is a separate folder
• Separation of content (raw prompts & configs) from presentation (static site)
• Integration with third-party services as loosely coupled “micro-services” for metrics & feedback
• Extendability: adding a new agent simply means adding a new folder with .txt & optional .json