Skip to content
Prompt

vibe-tools-mcp — Cursor Rules

by FutureExcited

AI Summary

Cursor Rules that integrate vibe-tools CLI into your development environment, enabling direct access to multiple AI models and code-aware assistance commands like planning and repository analysis without leaving your editor.

Install

Copy this and paste it into Claude Code, Cursor, or any AI assistant:

I want to add the "vibe-tools-mcp — Cursor Rules" prompt rules to my project.
Repository: https://github.com/FutureExcited/vibe-tools-mcp

Please read the repo to find the rules/prompt file, then:
1. Download it to the correct location (.cursorrules, .windsurfrules, .github/prompts/, or project root — based on the file type)
2. If there's an existing rules file, merge the new rules in rather than overwriting
3. Confirm what was added

Description

Cursor Rules for vibe-tools-mcp

Instructions

Use the following commands to get AI assistance: Direct Model Queries: vibe-tools ask "<your question>" --provider <provider> --model <model> - Ask any model from any provider a direct question (e.g., vibe-tools ask "What is the capital of France?" --provider openai --model o3-mini). Note that this command is generally less useful than other commands like repo or plan because it does not include any context from your codebase or repository. In general you should not use the ask command because it does not include any context. The other commands like web, doc, repo, or plan are usually better. If you are using it, make sure to include in your question all the information and context that the model might need to answer usefully. Ask Command Options: --provider=<provider>: AI provider to use (openai, anthropic, perplexity, gemini, modelbox, openrouter, or xai) --model=<model>: Model to use (required for the ask command) --reasoning-effort=<low|medium|high>: Control the depth of reasoning for supported models (OpenAI o1/o3-mini models and Claude 4 Sonnet). Higher values produce more thorough responses for complex questions. --with-doc=<doc_url>: Fetch content from one or more document URLs and include it as context. Can be specified multiple times (e.g., --with-doc=<url1> --with-doc=<url2>). Implementation Planning: vibe-tools plan "<query>" - Generate a focused implementation plan using AI (e.g., vibe-tools plan "Add user authentication to the login page") The plan command uses multiple AI models to: • Identify relevant files in your codebase (using Gemini by default) • Extract content from those files • Generate a detailed implementation plan (using OpenAI o3-mini by default) Plan Command Options: --fileProvider=<provider>: Provider for file identification (gemini, openai, anthropic, perplexity, modelbox, openrouter, or xai) --thinkingProvider=<provider>: Provider for plan generation (gemini, openai, anthropic, perplexity, modelbox, openrouter, or xai) --fileModel=<model>: Model to use for file identification --thinkingModel=<model>: Model to use for plan generation --with-doc=<doc_url>: Fetch content from one or more document URLs and include it as context for both file identification and planning. Can be specified multiple times (e.g., --with-doc=<url1> --with-doc=<url2>). Web Search: vibe-tools web "<your question>" - Get answers from the web using a provider that supports web search (e.g., Perplexity models and Gemini Models either directly or from OpenRouter or ModelBox) (e.g., vibe-tools web "latest shadcn/ui installation instructions") Note: web is a smart autonomous agent with access to the internet and an extensive up to date knowledge base. Web is NOT a web search engine. Always ask the agent for what you want using a proper sentence, do not just send it a list of keywords. In your question to web include the context and the goal that you're trying to acheive so that it can help you most effectively. when using web for complex queries suggest writing the output to a file somewhere like local-research/<query summary>.md. IMPORTANT: Do NOT use the web command for specific URLs. If a user provides a specific URL (documentation link, GitHub repo, article, etc.), you should always use commands that support the --with-doc parameter instead, such as repo, plan, doc, or ask. Using --with-doc ensures the exact content of the URL is processed correctly and completely. Web Command Options: --provider=<provider>: AI provider to use (perplexity, gemini, modelbox, or openrouter) Repository Context: vibe-tools repo "<your question>" [--subdir=<path>] [--from-github=<username/repo>] [--with-doc=<doc_url>...] - Get context-aware answers about this repository using Google Gemini (e.g., vibe-tools repo "explain authentication flow") Use the optional --subdir parameter to analyze a specific subdirectory instead of the entire repository (e.g., vibe-tools repo "explain the code structure" --subdir=src/components). Use the optional --from-github parameter to analyze a remote GitHub repository without cloning it locally (e.g., vibe-tools repo "explain the authentication system" --from-github=username/repo-name). Use the optional --with-doc parameter multiple times to include content from several URLs as additional context (e.g., vibe-tools repo "summarize findings" --with-doc=https://example.com/spec1 --with-doc=https://example.com/spec2). Documentation Generation: vibe-tools doc [options] [--with-doc=<doc_url>...] - Generate comprehensive documentation for this repository (e.g., vibe-tools doc --output docs.md). Can incorporate document context from multiple URLs (e.g., vibe-tools doc --with-doc=https://example.com/existing-docs --with-doc=https://example.com/new-spec). YouTube Video Analysis: vibe-tools youtube "<youtube-url>" [question] [--type=<summary|transcript|plan|review|custom>] - Analyze YouTube videos and generate detailed reports (e.g., vibe-tools youtube "https://youtu.be/43c-Sm5GMbc" --type=summary) Note: The YouTube command requires a GEMINI_API_KEY to be set in your environment or .vibe-tools.env file as the GEMINI API is the only interface that supports YouTube analysis. GitHub Information: vibe-tools github pr [number] - Get the last 10 PRs, or a specific PR by number (e.g., vibe-tools github pr 123) vibe-tools github issue [number] - Get the last 10 issues, or a specific issue by number (e.g., vibe-tools github issue 456) ClickUp Information: vibe-tools clickup task <task_id> - Get detailed information about a ClickUp task including description, comments, status, assignees, and metadata (e.g., vibe-tools clickup task "task_id") Wait Command: vibe-tools wait <seconds> - Pauses execution for the specified number of seconds (e.g., vibe-tools wait 5 to wait for 5 seconds). Model Context Protocol (MCP) Commands: Use the following commands to interact with MCP servers and their specialized tools: vibe-tools mcp search "<query>" - Search the MCP Marketplace and GitHub for available servers that match your needs (e.g., vibe-tools mcp search "git repository management") vibe-tools mcp run "<query>" - Execute MCP server tools using natural language queries (e.g., vibe-tools mcp run "list files in the current directory" --provider=openrouter). The query must include sufficient information for vibe-tools to determine which server to use, provide plenty of context. The search command helps you discover servers in the MCP Marketplace and on GitHub based on their capabilities and your requirements. The run command automatically selects and executes appropriate tools from these servers based on your natural language queries. If you want to use a specific server include the server name in your query. E.g. vibe-tools mcp run "using the mcp-server-sqlite list files in directory --provider=openrouter" Notes on MCP Commands: • MCP commands require ANTHROPIC_API_KEY or OPENROUTER_API_KEY to be set in your environment • By default the mcp command uses Anthropic, but takes a --provider argument that can be set to 'anthropic' or 'openrouter' • Results are streamed in real-time for immediate feedback • Tool calls are automatically cached to prevent redundant operations • Often the MCP server will not be able to run because environment variables are not set. If this happens ask the user to add the missing environment variables to the cursor tools env file at ~/.vibe-tools/.env Stagehand Browser Automation: vibe-tools browser open <url> [options] - Open a URL and capture page content, console logs, and network activity (e.g., vibe-tools browser open "https://example.com" --html) vibe-tools browser act "<instruction>" --url=<url | 'current'> [options] - Execute actions on a webpage using natural language instructions (e.g., vibe-tools browser act "Click Login" --url=https://example.com) vibe-tools browser observe "<instruction>" --url=<url> [options] - Observe interactive elements on a webpage and suggest possible actions (e.g., vibe-tools browser observe "interactive elements" --url=https://example.com) vibe-tools browser extract "<instruction>" --url=<url> [options] - Extract data from a webpage based on natural language instructions (e.g., vibe-tools browser extract "product names" --url=https://example.com/products) vibe-tools browser mac-chrome [options] - Start a Chrome instance with remote debugging (macOS only) (e.g., vibe-tools browser mac-chrome --debug, vibe-tools browser mac-chrome --lite) Notes on Browser Commands: • All browser commands are stateless unless --connect-to is used to connect to a long-lived interactive session. In disconnected mode each command starts with a fresh browser instance and closes it when done. • If you want to start a new long-lived session • When using --connect-to, special URL values are supported: • current: Use the existing page without reloading • reload-current: Use the existing page and refresh it (useful in development) • If working interactively with a user you should always use --url=current unless you specifically want to navigate to a different page. Setting the url to anything else will cause a page refresh loosing current state. • Multi step workflows involving state or combining multiple actions are supported in the act command using the pipe (|) separator (e.g., vibe-tools browser act "Click Login | Type 'user@example.com' into email | Click Submit" --url=https://example.com) • Video recording is available for all browser

Discussion

0/2000
Loading comments...

Health Signals

MaintenanceCommitted 10mo ago
Stale
AdoptionUnder 100 stars
5 ★ · Niche
DocsMissing or thin
Undocumented

GitHub Signals

Stars5
Issues0
Updated10mo ago
View on GitHub
MIT License

My Fox Den

Community Rating

Sign in to rate this booster

Works With

Any AI assistant that accepts custom rules or system prompts

Claude
ChatGPT
Cursor
Windsurf
Copilot
+ more