Create MCP servers for LLM interaction with external services.
npx skills add https://github.com/pacphi/sindri --skill mcp-builder-pacphi
Please help me install this Agent Skill. Skill: mcp-builder Source: https://github.com/pacphi/sindri/tree/main/docker/lib/extensions/vf-mcp-builder/resources Command: npx skills add https://github.com/pacphi/sindri --skill mcp-builder-pacphi
This Skill guides building MCP servers to enable LLMs to safely interact with external systems.
Create a simple MCP server skeleton and register a test tool.
High-intent search queries and answers about installing and using this skill.
Build an MCP server by registering tools with well-defined input/output schemas, choosing a transport (stdio, SSE, or HTTP), and implementing error handling. This Skill covers both Python (FastMCP) and Node/TypeScript implementations using Pydantic or Zod for validation, enabling safe LLM integration with external services.
Design tool schemas with clear, scoped inputs and outputs using structured validation (Pydantic for Python, Zod for TypeScript). Include descriptive naming, comprehensive error handling, and pagination support where needed. Follow MCP SDK best practices to ensure discoverability and safe operation.
Yes. This Skill applies to both Python (FastMCP) and Node/TypeScript MCP servers, covering tool registration, schema validation, transport choices, and security practices across both ecosystems.
Implement input validation using Pydantic (Python) or Zod (TypeScript) to enforce structured schemas before tool execution. Add comprehensive error handling and return multi-format responses that clearly communicate failures to the LLM.
Choose between stdio, streamable HTTP, or SSE (Server-Sent Events) depending on your integration context. Each transport has trade-offs for latency, scalability, and deployment complexity; this Skill covers selection criteria and implementation for each.
Apply security best practices including input validation, structured tool outputs, proper error handling to avoid exposing internals, and adherence to MCP SDK guidelines. This ensures robust integrations that safely expose external services to LLM access.
fritzprixDesigns and scaffolds MCP servers in Python FastMCP or Node/TypeScript SDKs with validation and error handling.
7a336e6eAutomate Flask deployment with Gunicorn, Docker, and Nginx configurations.
mtsatryanDesign, optimize, and evaluate LLM prompts across RAG workflows and benchmarks.
greghavensGuide building Triggerfish integrations using six patterns with TypeScript interfaces.
