io.github.bch1212/modelwatch

io.github.bch1212/modelwatch·v0.1.1·AI & LLMs

Quality Score

/100

Continuous behavioral drift monitoring for LLM apps — catches silent provider model updates.

Repository →Website →

§00 Trust verdict · v1 advisory · method

UNVERIFIEDno verdict on file

Verdict not yet evaluated for this tool. The hybrid eval runs adversarial cases first; coverage rolls out as the corpus expands. Until a verdict is recorded, an agent should treat this tool as not-yet-cleared and fall back to its own checks. Method: hybrid eval, four-state verdict, honest limits.

§01 Install

Claude Desktop (claude_desktop_config.json)

{
  "mcpServers": {
    "modelwatch": {
      "command": "npx",
      "args": [
        "-y",
        "modelwatch-mcp"
      ],
      "env": {
        "MODELWATCH_API_KEY": "<your-modelwatch_api_key>",
        "MODELWATCH_API_BASE": "https://api.modelwatch.app"
      }
    }
  }
}

Cursor (.cursor/mcp.json)

{
  "mcpServers": {
    "modelwatch": {
      "command": "npx",
      "args": [
        "-y",
        "modelwatch-mcp"
      ],
      "env": {
        "MODELWATCH_API_KEY": "<your-modelwatch_api_key>",
        "MODELWATCH_API_BASE": "https://api.modelwatch.app"
      }
    }
  }
}

Cline (cline_mcp_settings.json)