gog-weekly-digest
Pass
Audited by Gen Agent Trust Hub on Jun 24, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes the
gogcommand-line tool to fetch data. All commands include the--readonlyand--wrap-untrustedflags, and Gmail queries specifically use the--gmail-no-sendflag to prevent any state-changing actions. - [PROMPT_INJECTION]: The skill processes potentially malicious content from external sources (Gmail, Calendar, and Tasks). It mitigates indirect prompt injection risks by using
--wrap-untrustedmarkers and explicit instructions to the agent to avoid modifying data or sending communications. - [SAFE]: The ingestion of untrusted data is well-guarded: Ingestion occurs via the
gogtool search outputs (SKILL.md); Boundary markers are present via the--wrap-untrustedflag; Capabilities are restricted to read-only operations; Sanitization is handled by the tool's wrapping mechanism.
Audit Metadata