Lesson 365 of 1596
MCP Servers: Adding New Capabilities
Model Context Protocol turns any tool into something Claude Code can call. Adding the right MCP servers expands what the agent can actually do for you.
Creators · Tools Literacy · ~6 min read
Why MCP exists
Before MCP, every agent had its own bespoke way of integrating with every tool — different glue per agent, per service. MCP standardizes the contract: a server exposes tools, an agent calls them, both sides speak the same protocol. That standard turned 'plug Claude into Linear' from a custom integration into a one-line config.
What MCP gives Claude Code
- 1Tools beyond the filesystem and shell — calendar, ticketing, design, data, deploy
- 2Per-server permission scoping — Claude can only do what the server exposes
- 3A growing ecosystem of community servers for almost every product
- 4Symmetry with other agents — the same MCP servers work in Cursor, Windsurf, and beyond
- 5Local servers (process on your machine) and remote servers (HTTP)
Servers worth wiring up early
- A web search / browse server — for retrieval the agent can drive itself
- A ticketing / issue server (Linear, GitHub Issues, Jira) — agents that resolve tickets, not just discuss them
- A docs / wiki server — Notion, Confluence, your team's internal pages
- A scheduling server — read calendars, propose times, never auto-send invites without approval
- A data warehouse server (Postgres, Snowflake) read-only, with strict scopes
A typical MCP server config block. Local servers run as processes; remote servers connect via HTTP.
{ "mcpServers": { "linear": { "command": "npx", "args": ["-y", "@org/linear-mcp"], "env": { "LINEAR_API_KEY": "$LINEAR_API_KEY" } }, "docs": { "url": "https://mcp.example.com/docs", "transport": "http" } } }Apply: add one server thoughtfully
- 1Pick a tool you spend real time in (Linear, Notion, GitHub)
- 2Find the official or top-rated MCP server for it
- 3Read the source — confirm it does what it says, scoped how it claims
- 4Wire it up with the most restrictive token scope that still works
- 5Use it for a week; decide if it earned its keep before adding the next one
Key terms in this lesson
The big idea: MCP is how the agent gets bigger hands. Add servers deliberately, scope them tightly, and audit before installing.
End-of-lesson quiz
Check what stuck
8 questions · Score saves to your progress.
Tutor
Curious about “MCP Servers: Adding New Capabilities”?
Ask anything about this lesson. I’ll answer using just what you’re reading — short, friendly, grounded.
Progress saved locally in this browser. Sign in to sync across devices.
Related lessons
Keep going
Creators · 10 min
Building A Personal Research Stack With Perplexity At The Core
Perplexity is best as one tool in a stack. Here is how to combine it with reading apps, note tools, and primary-source databases for a workflow that compounds.
Creators · 10 min
Codex With Custom Tools And MCP
Codex's real power shows when you connect it to your own tools — internal APIs, datastores, ticketing systems — usually via Model Context Protocol.
Creators · 10 min
Skill Registries, Sharing, And Trust
Skills are code that runs in your soul's context. A registry is how you share them — and how attackers ship them. Public versus private registries, signing, permission scopes, and a security review checklist. OpenClaw maintainers and the broader local-agent community converge on a single warning: skills are the new supply-chain attack surface.