MCP 服务器

MCP Observatory

Finds problems in your MCP servers before your users do. Scans your Claude config, checks every server's capabilities, actually calls tools to make sure they work, and catches breaking changes between versions.

README

MCP Observatory

  ███╗   ███╗ ██████╗██████╗
  ████╗ ████║██╔════╝██╔══██╗
  ██╔████╔██║██║     ██████╔╝
  ██║╚██╔╝██║██║     ██╔═══╝
  ██║ ╚═╝ ██║╚██████╗██║
  ╚═╝     ╚═╝ ╚═════╝╚═╝
     O B S E R V A T O R Y

Find problems in your MCP servers before your users do.

You update a server, a tool silently breaks, and your agent starts failing. MCP Observatory catches that. It connects to your servers, checks every capability, actually calls tools to make sure they work, and diffs runs to catch what changed.

Quick Start

Scan every MCP server in your Claude config:

npx @kryptosai/mcp-observatory

Go deeper — also invoke safe tools to verify they actually run:

npx @kryptosai/mcp-observatory scan deep

Test a specific server:

npx @kryptosai/mcp-observatory test npx -y @modelcontextprotocol/server-everything

Add it to Claude Code as an MCP server:

claude mcp add mcp-observatory -- npx -y @kryptosai/mcp-observatory serve

Or add it manually to your config:

{
  "mcpServers": {
    "mcp-observatory": {
      "command": "npx",
      "args": ["-y", "@kryptosai/mcp-observatory", "serve"]
    }
  }
}

Commands

Command	What it does
`scan`	Auto-discover servers from config files and check them all (default)
`scan deep`	Scan and also invoke safe tools to verify they execute
`test <cmd>`	Test a specific server by command
`diff <base> <head>`	Compare two run artifacts for regressions and schema drift
`watch <config>`	Watch a server for changes, alert on regressions
`serve`	Start as an MCP server for AI agents

What It Does

Check capabilities — connects to a server and verifies tools, prompts, and resources respond correctly.

Invoke tools — goes beyond listing. Actually calls safe tools (no required params / readOnlyHint) and reports which ones work and which ones crash.

npx @kryptosai/mcp-observatory scan deep

Detect schema drift — diffs two runs and surfaces added/removed fields, type changes, and breaking parameter changes.

npx @kryptosai/mcp-observatory diff run-a.json run-b.json

Recommend servers — scans your project for languages, frameworks, databases, and cloud providers, then cross-references the MCP registry to suggest servers you're missing. Ask your agent "what MCP servers should I add?" and it figures out the rest.

Watch for regressions — re-runs checks on an interval and alerts when something changes.

npx @kryptosai/mcp-observatory watch target.json

Scan locations

When you run scan, it looks for MCP configs in:

~/.claude.json (Claude Code)
~/Library/Application Support/Claude/claude_desktop_config.json (Claude Desktop, macOS)
%APPDATA%/Claude/claude_desktop_config.json (Claude Desktop, Windows)
.claude.json and .mcp.json (current directory)

MCP Server Mode

When running as an MCP server (serve), your AI agent gets five tools:

Tool	What it does
`scan`	Discover and check all configured servers
`check_server`	Check a specific server by command
`diff_runs`	Compare two saved run artifacts
`get_last_run`	Return the most recent run for a target
`suggest_servers`	Scan your environment and recommend servers you're missing

An AI tool that checks other AI tools. It's a tool testing tools that serve tools.*

<sub>* I'm a dude playing a dude disguised as another dude.</sub>

Compatibility

Works with any MCP server that uses standard transports:

Transport	Examples	Adapter
stdio (most servers)	filesystem, memory, context7, brave-search, sentry, notion, stripe	`local-process`
HTTP/SSE (remote)	Cloudflare, Exa, Tavily	`http`
Docker	All `@modelcontextprotocol/server-*` images	`local-process` via `docker run -i`

Servers needing API keys work via env in the target config. Python servers work via uvx. See the full compatibility matrix for tested servers and known issues.

Target config files

For more control (env vars, metadata, custom timeout):

{
  "targetId": "filesystem-server",
  "adapter": "local-process",
  "command": "npx",
  "args": ["-y", "@modelcontextprotocol/server-filesystem", "."],
  "timeoutMs": 15000
}

npx @kryptosai/mcp-observatory run --target ./target.json

HTTP / SSE targets

{
  "targetId": "my-remote-server",
  "adapter": "http",
  "url": "http://localhost:3000/mcp",
  "authToken": "optional-bearer-token",
  "timeoutMs": 15000
}

Limitations

Servers requiring interactive OAuth (e.g., Google Drive) need pre-authentication before Observatory can connect
Custom WebSocket transports (e.g., BrowserTools MCP) are not supported
A few servers time out or close before init — see known issues and compatibility

Contributing

See CONTRIBUTING.md for guidelines. The fastest way to contribute is to add a real passing target with a distinct capability shape, a clearer report surface, or a cleaner startup diagnosis.