fauxshield_mcp 1.0.0

Model Context Protocol server for the FauxShield anti-counterfeit platform. Gives AI agents scoped, auditable access to products, batches, serialized units, fraud alerts, recalls, and the audit trail.

1.0.0

First published version of the FauxShield MCP server.

• Speaks the Model Context Protocol on stdin/stdout, so an MCP-speaking client (Claude Desktop, Cursor, …) can read the fraud queue, pull evidence, and — with a key that has write scopes — work alerts and recalls, always attributed in the audit log.
• 55 tools covering the platform's endpoints, from catalogs and verifications to alerts, recalls, analytics, and the audit trail.
• FAUXSHIELD_READ_ONLY=true hides every mutating tool.
• Configuration comes from the environment (FAUXSHIELD_URL, FAUXSHIELD_API_KEY or FAUXSHIELD_TOKEN, workspace, read-only); a misconfiguration is one readable line on stderr and a non-zero exit instead of an empty tool list.