MCP_INTEGRATION

A guide for connecting Station with MCP, a standard that lets AI systems discover and use external tools and services. Station can act both as a tool user and as a provider of tools to Claude Code or other clients.

In plain words
What is it for?
Use it to add MCP servers, expose Station's agent functions to Claude Code or Claude Desktop, discover available tools, and design multi-tool agent workflows.
Why use it?
It explains how to connect agents to tools through a consistent interface rather than writing a separate integration for every tool.

Agent

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add agents/cloudshipai/station/mcp_integration
Clone the repo
git clone --depth 1 https://github.com/cloudshipai/station
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 3,789 The whole file, excluding the scripts and references it only reads on demand.
Security scan C 2 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00000 $0.03789
Opus 5 $0.00000 $0.01895
Sonnet 5 $0.00000 $0.00758
Haiku 4.5 $0.00000 $0.00379

Measured 2d ago against content hash 7dd0dfd1afc7, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade C, and why

MCP_INTEGRATION scanned grade C with 2 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Downloads and executes remote codehighSupply chain

curl | sh runs whatever the server returns today, which is not necessarily what it returned when this was reviewed.

curl -sSL https://ship.sh/install | bash

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

curl -sSL https://ship.sh/install | bash
docs/agents/MCP_INTEGRATION.md · 646 lines

How it starts

The opening of the file, as written. The whole thing — 646 lines — stays where its author put it; the contents beside it link to each section on GitHub.

MCP Integration Guide

Station provides comprehensive integration with the Model Context Protocol (MCP), enabling agents to access powerful tools through standardized interfaces. This guide covers MCP server setup, tool discovery, and advanced integration patterns.

MCP Architecture in Station

MCP Protocol Overview

Station implements both MCP client and server capabilities:

┌─────────────────┐    ┌─────────────────┐    ┌─────────────────┐
│   Claude Code   │───▶│    Station      │───▶│   MCP Servers   │
│  (MCP Client)   │MCP │  (MCP Client    │MCP │  (Tools/APIs)   │
│                 │    │   & Server)     │    │                 │
└─────────────────┘    └─────────────────┘    └─────────────────┘
                              │
                       ┌──────▼───────┐
                       │  Station's   │
                       │ MCP Server   │
                       │ (13 tools)   │
                       └──────────────┘

Station's MCP Server

Station itself provides MCP tools via stdio interface:

Claude Code
claude mcp add --transport stdio station -e OTEL_EXPORTER_OTLP_ENDPOINT=http://localhost:4318 -- stn stdio
Claude Desktop

Edit ~/Library/Application Support/Claude/claude_desktop_config.json:

{
  "mcpServers": {
    "station": {
      "command": "stn",
      "args": ["stdio"],
      "env": {
        "OTEL_EXPORTER_OTLP_ENDPOINT": "http://localhost:4318"
      }
    }
  }
}
Cursor / Other MCP Clients

Add to .mcp.json in your project:

{
  "mcpServers": {
    "station": {
      "command": "stn",
      "args": ["stdio"],
      "env": {
        "OTEL_EXPORTER_OTLP_ENDPOINT": "http://localhost:4318"
      }
    }
  }
}

Note: The OTEL_EXPORTER_OTLP_ENDPOINT environment variable enables OpenTelemetry tracing. Station ships with Jaeger for trace visualization at http://localhost:16686.

Available Station Tools:

  • call_agent - Execute agents with custom parameters
  • create_agent - Create new agents with schemas
  • list_agents - Browse available agents
  • list_environments - View environment configurations
  • discover_tools - Find available MCP tools
  • create_bundle_from_environment - Package environments
  • sync_environment - Update MCP server connections

Read the full file on GitHub · 646 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 646 lines · 0 tokens per session scan C 7dd0dfd1afc7

Subscribe to this mod's changes

MCP_INTEGRATION is an agent published in the GitHub repository cloudshipai/station (429 stars, last pushed 7mo ago), licensed Apache-2.0. It costs nothing until one of its globs matches a file; then it loads 3,789 tokens. A static security scan graded it C with 2 findings (downloads and executes remote code, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other agents, from other repositories

demo-site

Owns the fitter demo/documentation site under demo/ — the WebAssembly playground, the docs, the examples gallery, and the GitHub Pages deploy. Use for any work on demo/index.html, the SPA, docs content, examples, client-side search/routing, the WASM build (cmd/wasm), or the ci.yaml pages job. Trigger phrases: "demo…

PxyUp/fitter · 97 tokens

deploy-reviewer

Reviews fold's deployment surface — the Helm chart's rendered output, Kubernetes security posture, the four Dockerfiles, and the compose stack — for what only shows up in the manifest rather than in the templates. Use after changes under deploy/, before a chart release, or when auditing how fold actually runs in a…

fold-run/fold · 67 tokens

release-verifier

Verifies a published fold release end to end from the outside — archives, checksums and their cosign signature, SBOMs, the four ghcr images, the OCI chart, and every sigstore attestation. Use after a tag's release workflow goes green, before announcing a release, or when someone asks whether a published artifact is…

fold-run/fold · 74 tokens

mcp-spec-auditor

Audits fold against the normative MCP specification for the pinned revision — method shapes, required fields and headers, error-code allocation, caching and intermediary rules, and deprecations on the clock. Use when a change touches wire behavior, when a new protocol revision ships, or before bumping the conformance…

fold-run/fold · 73 tokens

gateway-reviewer

Reviews fold changes against the gateway's architectural invariants — pipeline order, snapshot-based reload state, audit as the single exit door, the invisibility rule, the minted error-code registry, and proxy-path allocation discipline. Use proactively after any non-trivial change to gateway/, auth/, policy/, or…

fold-run/fold · 71 tokens

integration-test-author

Writes and extends fold's integration tests using real MCP SDK peers behind the gateway. Use when a change needs test coverage — new gateway behavior, reload/discovery paths, policy or auth changes, or a repro for a reported bug.

fold-run/fold · 50 tokens