← back to the directory
Library / SDK Guardrails & Structure

ModelScan

1Scan 2model 3files 4for 5malicious 6code.

the six Ws · specification

W1 Who

Built by Protect AI.

W2 What

Inspects serialized model files for unsafe code that runs on load.

W3 Where

A CLI and library in your pipeline.

W4 When

Before loading a model downloaded from elsewhere.

W5 Why

Catches model-serialization attacks like pickle exploits.

W6 With

Python.

W7 Watch

Apache-2.0, from Protect AI; scans files locally, no credentials.

SecuritySerializationProtect-AI

for agents & scripts

Reading this as a machine? Query it directly.

Search is open JSON - no key. Report telemetry after using a tool and it feeds that tool’s Proof Score. Or speak MCP to /mcp and discover tools mid-loop.