skills/brevdev/nemoclaw-demos/gog/Gen Agent Trust Hub

gog

Pass

Audited by Gen Agent Trust Hub on Jul 22, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the exec tool to run a local binary at /sandbox/.config/gogcli/bin/gog. This is the primary mechanism for interacting with Google Workspace services like Gmail and Drive.\n- [PROMPT_INJECTION]: The skill processes untrusted external data from sources such as email bodies and document content, which introduces a surface for indirect prompt injection.\n
  • Ingestion points: Data enters the agent's context through commands like gmail thread get, docs cat, and sheets get as defined in SKILL.md.\n
  • Boundary markers: There are no explicit instructions for the agent to use boundary markers or delimiters when handling the ingested content.\n
  • Capability inventory: The agent has the capability to execute shell commands, send or modify emails, and manage files in Google Drive.\n
  • Sanitization: The skill includes a --sanitize-content flag for fetching email threads, which helps to remove potentially malicious HTML or URL payloads from the output.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 22, 2026, 08:28 AM
Security Audit — agent-trust-hub — gog