bggg-data-x
Pass
Audited by Gen Agent Trust Hub on Aug 14, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill instructions in SKILL.md and references/chrome_collection.md explicitly forbid the extraction, inspection, or persistence of sensitive browser data, including cookies, local storage, passwords, and session tokens.
- [INDIRECT_PROMPT_INJECTION]: The skill handles untrusted data from the external web, creating a potential surface for indirect prompt injection. 1. Ingestion points: Extracted content from the X DOM via the browser collection loop in references/chrome_collection.md. 2. Boundary markers: Data is stored in structured JSON and JSONL formats as defined in references/schema.md. 3. Capability inventory: Local Python scripts (scripts/build_query_plan.py and scripts/normalize_x_dom.py) are used for data management and normalization. 4. Sanitization: The skill preserves raw text for auditability without applying sanitization or filtering to the external content.
Audit Metadata