uipath-project-handoff

Pass

Audited by Gen Agent Trust Hub on Jul 31, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The instructions explicitly include a 'Redact and bound' step, requiring the agent to remove credentials, tokens, personal data, and customer content before sharing the handoff.
  • [DATA_EXFILTRATION]: The skill aggregates project information like logs and design documents; it mitigates exfiltration risk by instructing the agent to provide pointers rather than duplicating sensitive content and by requiring redaction.
  • [COMMAND_EXECUTION]: The skill includes instructions to write output to a file and potentially initiate other agent sessions, but limits these actions to scenarios where the user provides explicit authorization.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 31, 2026, 04:38 PM
Security Audit — agent-trust-hub — uipath-project-handoff