gws-workflow-weekly-digest

Pass

Audited by Gen Agent Trust Hub on Apr 16, 2026

Risk Level: SAFE
Full Analysis
  • [DATA_EXPOSURE]: The skill is designed to access sensitive user data, including calendar meeting details and email metadata (unread counts). This behavior is transparently documented and aligns with the skill's stated purpose of providing a productivity digest.
  • [INDIRECT_PROMPT_INJECTION]: Because the skill processes external data (meeting titles and email summaries), it presents an indirect prompt injection surface. Malicious content within a calendar invite or an email could potentially attempt to influence the agent's behavior during the generation of the digest. No boundary markers or specific sanitization steps are explicitly mentioned in the provided file.
  • [COMMAND_EXECUTION]: The skill utilizes a local command-line utility named gws. While the instruction set describes basic usage with flags like --format, the agent must ensure that any user-provided inputs used to construct these commands are properly sanitized to prevent local command injection.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 16, 2026, 03:09 PM