codeintegrity-ai
Endpoint security for AI agents: inspect tool and skill access, enforce configured runtime policy, and publish prompt-injection research.
Built from codeintegrity-ai's public GitHub profile, United States of America. Not on Take My Repo yet, so there are no posts or updates here.
CodeIntegrity swarm poisoning proof of concept: forged subagent output led a coding agent to install an attacker-controlled package.
CodeIntegrity DERAIL: adaptive prompt-injection research for tool-using agents and LLM-as-a-judge defenses.