SpiderFoot — Security Hardening (fork)

A fork of SpiderFoot, an open-source OSINT automation framework. To be precise about what is mine: the AI layer, meaning the analysis agents, the model gateway, the vector search and the event architecture, is the upstream maintainer's work. My contribution is a seven-commit security-hardening track. The piece I care most about is a guard against LLM data exfiltration. Once you feed scan results into a language model, the model's output becomes a route by which sensitive findings can leave the system, and it is a route people rarely think to close. The guard fails closed, so an error stops the flow rather than quietly letting it through.