Skills
skills/aaaaqwq/agi-super-skills/feishu-doc-optimizer/Gen Agent Trust Hub

feishu-doc-optimizer

Pass

Audited by Gen Agent Trust Hub on Mar 9, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection from external document content. \n
  • Ingestion points: scripts/feishu_doc_reader.py reads raw document content via API. \n
  • Boundary markers: None; the skill does not use delimiters to isolate document data from instructions. \n
  • Capability inventory: scripts/feishu_doc_editor.py uses Playwright to wipe and rewrite document content. \n
  • Sanitization: None; document content is processed directly. \n- [COMMAND_EXECUTION]: Automated browser interactions perform destructive operations on user documents. \n
  • Evidence: scripts/feishu_doc_editor.py contains a clear_document function that wipes all text by looping automated 'Control+a' and 'Backspace' keystrokes. \n- [DATA_EXFILTRATION]: Accesses and reads sensitive document content. \n
  • Evidence: scripts/feishu_doc_reader.py fetches the full raw content of documents via the Feishu Docx API for processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 9, 2026, 10:14 PM