Tool · curated 16 Aug 2026
GitHub - delphisecurity/xaidr
First reported github.com
Coverage timeline
Single-source research — first reported, latest, and curated coincide.
Why it matters
xaidr gives defenders a runnable sensor to detect and optionally block agentic-AI abuse patterns — role forgery, forged tool results, and credential exfiltration — within LLM tool-calling workflows.
xaidr (by delphisecurity) is an open-source, Apache-2.0 licensed detection sensor for AI-agent security that classifies and can enforce policy against threats such as part-level role forgery, forged tool-result injection, and credential/secret egress in agentic tool-call paths. The repository includes rules, tests, and configurable enforcement modes (classify-only vs. block) for actions like reads from managed secret stores.