gws-meet
Pass
Audited by Gen Agent Trust Hub on Mar 20, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilize the 'gws' command-line binary to perform Google Meet management tasks. This is a standard method for CLI-based agent skills.
- [PROMPT_INJECTION]: The skill ingests data from Google Meet transcripts and participant names, which presents a surface for indirect prompt injection. Ingestion points: Conference records and transcripts (SKILL.md). Boundary markers: No delimiters or ignore instructions for embedded content are present. Capability inventory: The agent can invoke 'gws' for read and write operations on meeting data. Sanitization: No evidence of validation or sanitization of ingested content is defined in the skill.
Audit Metadata