Apps

Prompts written for any general-purpose AI chat app, such as ChatGPT, Claude, Gemini, or Le Chat. They need no special tooling: paste the prompt into a new conversation and run it as-is.

Prompts built for Cursor, the AI code editor. They assume codebase-aware chat or agent mode inside the editor and are written to steer edits, refactors, and reviews over an open project.

Prompts built for GitHub Copilot's chat and agent modes in VS Code, JetBrains IDEs, and github.com. They assume workspace context and are written to direct edits, reviews, and coding agent tasks.

Prompts built for Google Antigravity, the agent-first development environment that absorbed Gemini CLI. They are written for autonomous agents that plan, edit, and verify work across the editor, terminal, and browser.

Prompts built for OpenClaw, the self-hosted personal AI agent reachable over WhatsApp, Telegram, Slack, and other messengers. They are written as personas, standing instructions, or tasks for an always-on assistant.

Prompts built for Cline, the open-source agent extension for VS Code. They assume editor and terminal access with human-in-the-loop approval of file changes and commands.

Prompts built for Devin, Cognition's autonomous software engineer, including Devin Desktop (formerly Windsurf). They are written as delegated tickets: a goal, constraints, and acceptance criteria for an agent that plans and executes end to end.

Prompts built for Junie, the JetBrains coding agent in IntelliJ-family IDEs and the Junie CLI. They assume full project context and are written to drive multi-step tasks inside JetBrains tooling.

Prompts built for Replit Agent, which builds and deploys full applications from natural-language briefs. They describe the product to build - features, data, and integrations - rather than individual code edits.

Prompts built for Lovable and similar vibe-coding app builders. They describe an application or feature in product terms and rely on the platform to generate, host, and iterate on the code.

Prompts built for Perplexity and other answer engines. They are research-shaped: questions, comparisons, and briefs that lean on live web search and citations.