gws-meet
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill relies on the execution of the
gwsbinary to interact with Google Workspace APIs. This is the intended functionality as specified in the metadata. - [INDIRECT_PROMPT_INJECTION]: The skill provides methods to retrieve untrusted external content that could contain malicious instructions.
- Ingestion points: Meeting metadata, conference records, participant lists, recordings, smart notes, and transcripts are ingested from the Google Meet API into the agent context via the
conferenceRecordsresource (SKILL.md). - Boundary markers: There are no explicit instructions providing boundary markers or delimiters to protect the agent from instructions embedded within retrieved transcripts or notes.
- Capability inventory: The agent has the capability to interact with the broader Google Workspace environment using the
gwstool and its associated resources. - Sanitization: The skill does not define any sanitization or validation logic for the content retrieved from meeting transcripts or notes.
Audit Metadata