@tryfinch/finch-api-mcp@10.8.0
The official MCP Server for the Finch API
AI security review
scanned · by lpm-firewall-aiSource inspection disproves the scanner's malicious conclusion: the risky primitives implement the documented MCP code-execution feature and do not establish a covert attack chain. The exposed code-execution capability still warrants a warning because MCP callers can invoke it with the server's Finch credentials and environment.