nerdearla
Mis demos en Nerdearla Spain 2025
claude mcp add --transport stdio 0gis0-nerdearla-mcp npx -y 0gis0-nerdearla-mcp
How to use
This MCP server bundle provides three demonstrations of MCP-powered tools centered around YouTube content interaction and generative capabilities. The tools expose an MCP-compliant interface that lets you search for videos, engage in interactive elicitations to refine user intent, and generate creative video titles using model-assisted sampling. Specifically, you can access:
- Search Video: a basic tool that retrieves YouTube videos matching a query. It demonstrates input validation, external API calls, robust error handling, and returning MCP-formatted results.
- Search Channel: an interactive tool that finds YouTube channels and engages the user through elicitation prompts to decide whether to fetch the channel’s latest videos. This showcases bidirectional dialogue, dynamic schema validation, and client-server communication features.
- Generate Video Title: an advanced creative tool that delegates title generation to the model, guiding the user through language selection and producing multiple creative title variants via sampling. These tools illustrate how to structure MCP tool definitions, validate inputs with schemas, call external services, and compose responses suitable for downstream MCP clients. To experiment, initialize the server and invoke each tool through your MCP client, observing the validation, dialog flows, and multi-result outputs that the server returns.
How to install
Prerequisites:
- Node.js (LTS) and npm installed on your machine
- Git (optional, for cloning the repo)
Install and run:
- Clone or download this MCP server repository to your machine.
- Install dependencies (from the repository root):
npm install
- Build if a build step is required (consult package.json scripts). If there is a TypeScript source, you may need:
npm run build
- Run the MCP server:
npm start
- Connect an MCP client to the running server endpoint and begin invoking the tools: Search Video, Search Channel, and Generate Video Title. If using npx as configured in mcp_config, the server will be launched via the package name when needed.
Notes:
- If the repository uses a different startup script, replace npm start with the exact command shown in the package.json scripts section (for example, node dist/server.js).
- Ensure any required API keys (e.g., YouTube Data API) are provided as environment variables as the server’s tools implement external service calls.
Additional notes
Tips and caveats:
- YouTube API usage may require an API key. If the tools reference the API, set the key in an environment variable (e.g., YT_API_KEY) or as documented by the repository.
- The Search Channel tool uses elicitation flows; be prepared to respond to prompts that refine what content you want (e.g., channel filters, latest videos).
- Tool outputs are structured to fit MCP response schemas; when integrating with clients, ensure your client handles validation errors and dialog state transitions gracefully.
- If you encounter schema validation errors, verify that inputs conform to the expected shapes defined in the tool schemas (e.g., required fields for query or selection enums).
- If running in a containerized environment, you may prefer the docker command form or adapt the npm start script to your orchestration workflow.
Related MCP Servers
obsidian -tools
Add Obsidian integrations like semantic search and custom Templater prompts to Claude or any MCP client.
ironcurtain
A secure* runtime for autonomous AI agents. Policy from plain-English constitutions. (*https://ironcurtain.dev)
create -app
A CLI tool for quickly scaffolding Model Context Protocol (MCP) server applications with TypeScript support and modern development tooling
awesome s
A curated list of excellent Model Context Protocol (MCP) servers.
lightdash
A MCP(Model Context Protocol) server that accesses to Lightdash
grok-faf
First MCP server for Grok | FAST⚡️AF • URL-based AI context • Vercel-deployed