Portfolio MCP Server
@amitshrivastava/portfolio-mcp is a real working MCP server (Claude Desktop, Cursor, Cline, …) that lets any AI agent search and read my blog posts via the Model Context Protocol. ~250 lines of TypeScript on top of @modelcontextprotocol/sdk, stdio transport, JSON-RPC 2.0. Exposes every published blog post as a `portfolio://blog/<slug>` resource, a `search_blog` tool with keyword matching, a `list_blog` tool, and a `discuss-amits-thinking-on` prompt that orchestrates search-then-summarize. It's the companion to the MCP Inspector demo in AI Labs — that demo shows what MCP looks like on the wire; this is the server that produces that traffic.
Built in public to learn something or prove a pattern. Some are published packages, some are demos that live on this site.
- ▸Exposing portfolio content to any AI client via a single open protocol, not custom APIs per client
- ▸Keeping the server lean and shippable — no embedding model, no database, no runtime config
- ▸Making it actually useful in real AI clients with minimal install friction
- ▸Implements the official MCP TypeScript SDK with stdio transport and JSON-RPC 2.0
- ▸Reads blog markdown at startup and ranks results with simple keyword matching — no database needed
- ▸Single config block for Claude Desktop; runs via `node dist/index.js` after a one-time npm install + build
Installable MCP server that any compatible AI client can use today
30+ blog posts exposed as MCP resources, plus tools and prompts
Demonstrates protocol-level AI integration, not just consuming someone else's API
JobJam.io — AI Job Search & Application Platform
AI-powered job search platform: discover roles, evaluate fit, tailor applications, and close skill gaps. One-time pricing, no subscriptions.
Fathohm — Comprehension-Debt System of Record
See how much of your codebase no human understands. A treemap of your repo colored by human comprehension score, with a tracked, assignable metric for 'who understands what' in an AI-native codebase.
Grenz — Firewall for AI Agents
A local firewall for AI agents: scoped, revocable, observable permissions instead of your full GitHub token. Deny-by-default, MIT licensed, live at grenz.dev.