doctorg
Warn
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- COMMAND_EXECUTION (MEDIUM): The skill instructions direct the agent to execute local Python scripts (health_query.py) to query a health database. This involves running shell commands with arguments that vary based on the topic.
- DATA_EXFILTRATION (MEDIUM): The skill systematically accesses highly sensitive personal information, including vitals, heart rate, sleep patterns, and workout history from a local database (health.db). This data is processed and included in the output, creating a high-risk surface for exposure of sensitive health data.
- PROMPT_INJECTION (LOW): The skill is vulnerable to Indirect Prompt Injection (Category 8) by ingesting full-text content via Firecrawl without sanitization. Evidence Chain: 1. Ingestion points: WebSearch, Tavily, Firecrawl. 2. Boundary markers: Absent. 3. Capability inventory: Local script execution (health_query.py) and local file access. 4. Sanitization: Absent.
Audit Metadata