English Français العربية

AgentBreakWatch

Official registry Security

Audit AI Agent SDKs and MCP servers for breaking changes, schema drift, and migration fixes.

Source code Website Documentation

What can you do with AgentBreakWatch?

  • check_stack_drift: Pass the raw contents of a .mcp.json, package.json, or requirements.txt file to detect breaking major version jumps, unpinned @latest MCP servers, and archived packages, with exact version pin commands.
  • get_package_stability: Retrieve live npm/PyPI version, publish date, 90-day release velocity, 0-100 stability score, tools/list schema drift risk, and copy-paste before/after migration code for any of the 256 tracked packages.
  • list_volatile_packages: Query the fastest-changing Agent SDKs and MCP servers over the last 90 days, optionally filtered by kind (sdk or mcp) or ecosystem category.
  • get_archived_mcp_replacements: Retrieve the migration matrix for archived @modelcontextprotocol/server-* packages (server-github, server-postgres, server-puppeteer, server-brave-search, server-redis) and their official 2026 vendor-maintain...

Install

Add this to your client's MCP configuration (Claude Desktop: claude_desktop_config.json; Cursor: .cursor/mcp.json; VS Code and others use the same "mcpServers" shape).

Remote (streamable-http)
Claude Code 
claude mcp add --transport http agentbreakwatch 'https://agentbreakwatch.com/api/mcp'
JSON configuration (Claude Desktop, Cursor, VS Code) 
{
    "mcpServers": {
        "agentbreakwatch": {
            "type": "http",
            "url": "https://agentbreakwatch.com/api/mcp"
        }
    }
}

Documentation

From the project's README on GitHub, refreshed weekly. Links open on GitHub.

Contents
  1. AgentBreakWatch MCP Server (agentbreakwatch-mcp)
  2. Quick Connect (Zero-Install Remote MCP)
  3. Available MCP Tools
  4. Browse Ecosystem Stability Hubs
  5. License

AgentBreakWatch MCP Server (agentbreakwatch-mcp)

Official Model Context Protocol (MCP) server for AgentBreakWatch, the live breaking-change and tools/list JSON Schema drift telemetry index across 30 core AI Agent SDKs and 226 MCP servers (256 total tracked packages) on npm and PyPI.

Quick Connect (Zero-Install Remote MCP)

Add the hosted Streamable HTTP endpoint directly to your .mcp.json, Cursor, Claude Code, or Claude Desktop config with zero local dependencies:

{
  "mcpServers": {
    "agentbreakwatch": {
      "url": "https://agentbreakwatch.com/api/mcp"
    }
  }
}

Or run locally via npx (stdio transport):

{
  "mcpServers": {
    "agentbreakwatch": {
      "command": "npx",
      "args": ["-y", "agentbreakwatch-mcp"]
    }
  }
}

Available MCP Tools

  1. check_stack_drift: Pass the raw contents of a .mcp.json, package.json, or requirements.txt file to detect breaking major version jumps, unpinned @latest MCP servers, and archived packages, with exact version pin commands.
  2. get_package_stability: Retrieve live npm/PyPI version, publish date, 90-day release velocity, 0-100 stability score, tools/list schema drift risk, and copy-paste before/after migration code for any of the 256 tracked packages.
  3. list_volatile_packages: Query the fastest-changing Agent SDKs and MCP servers over the last 90 days, optionally filtered by kind (sdk or mcp) or ecosystem category.
  4. get_archived_mcp_replacements: Retrieve the migration matrix for archived @modelcontextprotocol/server-* packages (server-github, server-postgres, server-puppeteer, server-brave-search, server-redis) and their official 2026 vendor-maintained replacements.

Browse Ecosystem Stability Hubs

License

MIT - AgentBreakWatch

Frequently asked questions

What is AgentBreakWatch?

Audit AI Agent SDKs and MCP servers for breaking changes, schema drift, and migration fixes.

How do I install AgentBreakWatch?

It runs as a remote server. In Claude Code: claude mcp add --transport http agentbreakwatch 'https://agentbreakwatch.com/api/mcp'. For other clients, paste the JSON configuration from this page into their MCP settings.

Does AgentBreakWatch need an API key?

Its registry entry lists no required key or secret. Some features may still need an account with the service it connects to.

Is AgentBreakWatch open source?

Yes. Its code is on GitHub under the MIT license.

Who maintains AgentBreakWatch?

It is published by joeyg6393. AI.tn lists it but does not maintain or endorse it.

From the official MCP Registry. Listing is not an endorsement: review a server's code and permissions before giving it access to your data.