Explore AI system prompts, agents, and tools
The largest open collection of real AI system prompts — from Anthropic Claude, Google Gemini, Cursor, GitHub Copilot, and more. Browse, compare, and install.
- Prompt Collections
- 32
- Total Files
- 100
- Total Lines
- 36,617
- Installable Agents
- 144
What are AI system prompts?
System prompts are the hidden instructions that shape how AI models behave. Every time you use ChatGPT, Claude, Gemini, or a coding assistant like Cursor, a system prompt is running behind the scenes — defining the model's personality, capabilities, and guardrails.
Studying real system prompt examples is one of the fastest ways to improve at prompt engineering. By comparing how Anthropic, Google, and other companies configure their AI products, you can learn patterns for writing better instructions, setting constraints, and getting more reliable outputs from any language model.
System Prompts
Explore prompts from Anthropic Claude, Google Gemini, Cursor, Copilot, and 28+ more
System prompts and configuration for Anthropic. Includes agent prompts and tool definitions. 4,579 lines across 7 files.
System prompts and configuration for Amp. 4,196 lines across 3 files.
System prompts and configuration for Open Source prompts. Contains agent/system prompts. 2,497 lines across 7 files.
System prompts and configuration for Google. Contains agent/system prompts. 2,471 lines across 3 files.
System prompts and configuration for Cursor Prompts. Includes agent prompts and tool definitions. 2,310 lines across 7 files.
System prompts and configuration for Leap.new. Includes agent prompts and tool definitions. 1,754 lines across 2 files.
System prompts and configuration for VSCode Agent. Contains agent/system prompts. 1,547 lines across 9 files.
System prompts and configuration for Augment Code. Includes agent prompts and tool definitions. 1,467 lines across 4 files.
AI Agents
144 agents across 12 divisions — install into Cursor, Claude Code, Windsurf, and more
Testing
Expert accessibility specialist who audits interfaces against WCAG standards, tests with assistive technologies, and ensures inclusive design. Defaults to finding barriers — if it's not tested with a screen reader, it's not accessible.
Sales
Expert post-sale account strategist specializing in land-and-expand execution, stakeholder mapping, QBR facilitation, and net revenue retention. Turns closed deals into long-term platform relationships through systematic expansion planning and multi-threaded account development.
Specialized
Autonomous payment processing specialist that executes vendor payments, contractor invoices, and recurring bills across any payment rail — crypto, fiat, stablecoins. Integrates with AI agent workflows via tool calls.
Paid Media
Paid media creative specialist focused on ad copywriting, RSA optimization, asset group design, and creative testing frameworks across Google, Meta, Microsoft, and programmatic platforms. Bridges the gap between performance data and persuasive messaging.
Engineering
Specialist in self-healing data pipelines — uses air-gapped local SLMs and semantic clustering to automatically detect, classify, and fix data anomalies at scale. Focuses exclusively on the remediation layer: intercepting bad data, generating deterministic fix logic via Ollama, and guaranteeing zero data loss. Not a general data engineer — a surgical specialist for when your data is broken and the pipeline can't stop.
Support
Expert data analyst transforming raw data into actionable business insights. Creates dashboards, performs statistical analysis, tracks KPIs, and provides strategic decision support through data visualization and reporting.
Marketing
Expert app store marketing specialist focused on App Store Optimization (ASO), conversion rate optimization, and app discoverability
Product
Behavioral psychology specialist that adapts software interaction cadences and styles to maximize user motivation and success.
How to use this directory
- 1
Browse or search the directory
Use the search bar or browse by category to find system prompts, agents, or tools relevant to your work.
- 2
Read and compare prompts
Open any prompt collection to read the full system prompt text, compare approaches across different AI tools, and study prompt engineering patterns.
- 3
Install or copy what you need
Copy prompt text for reference, or use the one-click install snippets to add agents directly into Cursor, Claude Code, Windsurf, and other tools.
Tools & Resources
A swarm intelligence prediction engine that builds parallel digital worlds with thousands of AI agents to simulate and predict future outcomes.
Fully automatic censorship removal for language models using directional ablation with TPE-based parameter optimization.
10 guides covering installation, onboarding, multi-agent routing, sandboxing, Docker, and VPS deployments.
Frequently asked questions
What is a system prompt?
A system prompt is a set of instructions given to an AI model before a conversation begins. It defines the model's behavior, personality, capabilities, and constraints. System prompts are how companies like Anthropic, Google, and OpenAI shape their AI products.
How do I install an AI agent into Cursor?
Find an agent on this directory, click the "Cursor" install tab, and copy the provided JSON snippet into your .cursor/agents/ directory. Each agent page includes ready-to-use install snippets for Cursor, Claude Code, Windsurf, and other supported tools.
Are these system prompts official?
The system prompts in this directory are collected from publicly available sources — open-source repositories, published documentation, and community contributions. They reflect real-world configurations but may not always be the latest production versions.
What coding tools are supported for agent installation?
Agents can be installed into Cursor, Claude Code, Windsurf, GitHub Copilot, Gemini CLI, Aider, OpenCode, and several other AI-powered development tools. Each agent page shows which tools are supported.
Is this directory open source?
Yes. The source code and all prompt data are publicly available on GitHub. Contributions, corrections, and new prompt submissions are welcome.