Tool
Maintaining Privacy With Agents: What Actually Works When Sensitive Data Is Part of the Workflow
Publication date not yet evaluated · Added jackdavis.net
Page published
Coverage timeline
Single-source research — one report is available.
Why it matters
Defenders building or deploying tool-using AI agents need control points to prevent sensitive data leaking into model context, and this provides a runnable approach to mixed-sensitivity information-flow control.
A write-up releasing `agent-privacy`, an open-source privacy-control layer for CLI-style agent harnesses that intercept sensitive data (PII, identifiers) before it becomes model-visible context. It uses prompt/pre-tool/post-tool hooks with four actions — allow, redact, handoff, block — and integrates a local filter service, OpenAI Privacy Filter, and local Qwen models, with an operational pii-guard report covering ~6,343 screening decisions.