Mcp

A fresh Claude Code install is useful. A configured one is a different tool entirely. The gap between the two is maybe fifteen minutes of setup — a code index, a handful of MCP servers, two plugin marketplaces, a spec workflow, and a runtime to keep it all alive — and I keep re-doing it every time I move to a new machine. So here it is written down, in the order I run it.

Every Claude Code session that involves more than one repo starts with the same monologue: “we have a payment service that calls user-service via REST, which publishes events that notification-service consumes, and oh by the way the platform team owns auth-service…” By the time I’ve finished the briefing, I’ve burned tokens and patience.

cartographer-mcp is my attempt at fixing that. It crawls your GitLab organization, builds a service catalog with a dependency graph, and exposes it as MCP tools so Claude Code already knows the shape of your architecture.

I spend a lot of time in Claude Code, and a lot of time in GitLab. The constant tab-switching between the two — copy an issue ID, paste it into a prompt, jump back to read a job log, paste that too — got tiresome. So I wrote gitlab-mcp, a Model Context Protocol server that lets Claude Code talk to GitLab directly.

A surprising amount of debugging boils down to “let me check the database.” Schema lookups, sample row counts, “is this index actually being used,” EXPLAIN plans. None of that is hard to do in psql, but pulling Claude Code into the loop means copy-pasting schemas and query results back and forth, which gets old.

postgresql-mcp is an MCP server that gives Claude Code direct, read-only access to PostgreSQL. Claude does the inspection, you keep the conversation flowing.

I maintain a Go library for the Perplexity API (perplexity-go). Once it existed, it was almost a sin not to wrap it in a CLI. So pplx is what I reach for when I want a cited, web-aware answer from the shell — and now also a Model Context Protocol server I plug into Claude Code.