skills/joelhooks/joelclaw/gogcli/Gen Agent Trust Hub

gogcli

Pass

Audited by Gen Agent Trust Hub on Feb 27, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes a local binary named gog and a credential management tool secrets to perform its operations. These commands are used to manage authentication and interact with Google Workspace APIs.
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection due to the way it processes external data.
  • Ingestion points: The agent can read untrusted content from email bodies (gog gmail messages search --include-body), Google Docs (gog docs cat), and Google Sheets (gog sheets get).
  • Boundary markers: There are no boundary markers or specific instructions defined in the skill to prevent the agent from obeying instructions embedded within the retrieved Google Workspace data.
  • Capability inventory: The skill provides powerful capabilities including sending emails (gog gmail send), uploading files to Drive (gog drive upload), and modifying calendar events or tasks.
  • Sanitization: No evidence of data sanitization or validation is present in the skill definition to filter malicious instructions from ingested content.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 27, 2026, 06:43 PM