@codingthirty: I learned how to build MCPs from @kentcdodds, and it's a gift that keeps on giving. I built a local-first knowledge bas…
Summary
Developer shares experience building a local-first knowledge base using MCPs, Strapi, TanStack, and Ollama with Gemma 4, noting easy switch to frontier models like Claude.
View Cached Full Text
Cached at: 04/21/26, 01:02 PM
I learned how to build MCPs from @kentcdodds, and it’s a gift that keeps on giving. I built a local-first knowledge base using @strapijs, @tan_stack, and @ollama with Gemma 4. But if you need to use a frontier model like Claude for more power, just expose your tools and
Similar Articles
I made a local-first MCP tutorial repo with node-llama-cpp and a custom agent loop
A tutorial repo that teaches the Model Context Protocol (MCP) by building it step by step in Node.js, with a local-first setup using node-llama-cpp, GGUF models, and a custom agent loop.
Built a production RAG chatbot with custom MCP servers as the action layer, sharing what I learned
A full-stack engineer shares learnings from building a production RAG chatbot with custom MCP servers as an action layer, enabling natural language triggers for live workflows.
@kentcdodds: MCP
Tyler Benfield discusses the coolness of MCP for giving AI agents data access, sharing his use case of Claude routines for infrastructure monitoring and improvement.
I built a code knowledge graph tool that's actually MIT licensed (fully local, no cloud)
An indie dev released repopedia, an MIT-licensed, fully local code knowledge-graph tool that builds a SQLite graph from Python/TypeScript repos using tree-sitter, plus an MCP server so coding agents like Claude Code can query call graphs and blast-radius impact directly instead of relying on grep.
@kentcdodds: Read source outside your stack - MacPaint, Linux, MediaWiki. @grady_booch on why curiosity across domains builds judgme…
Kent C. Dodds shares advice from Grady Booch on reading source code outside your tech stack (e.g., MacPaint, Linux, MediaWiki) to build unique judgment through cross-domain curiosity.