Provide an open test corpus to assess and improve AI agent egress security through validated cases and automated workflows.
bilkulsahi1235/agent-egress-bench is an MCP server distributed on github, maintained by bilkulsahi1235, tracked here since June 2026. It has shipped 0 releases. Its composite risk grade is B — an inferred review prompt computed from observed signals, not a verdict.
Reading the source raised one review prompt — dangerous code — each a pattern worth a human look rather than a finding of fault.
full security breakdown →No tools recorded for this server yet.
No releases yet.
No releases yet.