Skip to content

MCP

MCP stands for Model Context Protocol. It is an open way for agents to connect to external capabilities such as documentation sources, databases, internal tools, issue trackers, or custom systems.

Use MCP when the agent needs structured, repeatable access to a system. Do not use it just because it sounds advanced.

Without MCP, you often paste context into prompts or rely on generic web browsing. With MCP, the agent can get structured access to the exact system it needs.

That usually means:

  • less prompt stuffing
  • more reliable context retrieval
  • better tool discoverability
  • workflows that are easier to share across teams

MCP is a good fit when the agent needs repeatable access to:

  • official documentation
  • internal knowledge systems
  • tickets and project data
  • controlled tools that should not be re-described in every prompt
PieceWhat it does
Toolsfunctions the agent can call
Resourcesdata the agent can read as context
Promptsreusable prompt templates from the server
Serversthe processes exposing those capabilities
Server typeWhy people add it
GitHubissues, PRs, code search, repo actions
Browser automationscreenshots and UI verification
Filesystemcarefully scoped file access
Databaseread-only inspection and debugging
Web searchcurrent docs and release changes
  • Only connect servers you trust.
  • Assume external content can contain prompt injection.
  • Give the agent access to the smallest useful surface area.
  • Prefer read-only integrations unless write access is clearly justified.

If you keep pasting the same docs or instructions into prompts, that is a sign the information probably wants to live in a context file, a skill, or an MCP server instead.

{
"mcpServers": {
"github": {
"command": "npx",
"args": ["-y", "@modelcontextprotocol/server-github"],
"env": {
"GITHUB_TOKEN": "${GITHUB_TOKEN}"
}
},
"playwright": {
"command": "npx",
"args": ["-y", "@playwright/mcp-server"]
}
}
}

Every server adds more capability, but also more surface area and usually more context overhead. Keep only the servers that help with the task at hand.