Skills
skills/atomlink-ye/my-claude-plugins/google-workspace/Gen Agent Trust Hub

google-workspace

Pass

Audited by Gen Agent Trust Hub on May 13, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: Indirect prompt injection vulnerability surface detected.\n- Ingestion point: Document content retrieved via gws docs documents get in SKILL.md.\n- Boundary markers: No explicit markers or instructions to isolate document content from instructions.\n- Capability inventory: Subprocess execution of gws for reading and writing documents across Google Workspace.\n- Sanitization: scripts/extract-doc-text.py extracts raw text without filtering for instruction-like patterns.\n- Note: This finding is considered a risk surface inherent to the skill's primary document processing function.
Audit Metadata
Risk Level
SAFE
Analyzed
May 13, 2026, 03:05 AM