emergency-card
Pass
Audited by Gen Agent Trust Hub on Nov 3, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to run a local Python script (
python scripts/generate_emergency_card.py) to process the aggregated medical data and produce formatted output. This execution is part of the skill's primary intended functionality. - [INDIRECT_PROMPT_INJECTION]: The skill processes data from various local JSON files including user profiles, allergy lists, medication records, and surgical history. This ingestion represents an attack surface where malicious instructions embedded in health records could potentially influence the agent's behavior during the card generation process.
- Ingestion points: Multiple files including
data/profile.json,data/allergies.json,data/medications/medications.json, and glob patterns for surgical and discharge records. - Boundary markers: The instructions do not specify any delimiters or warnings to the agent to ignore instructions embedded within the processed data files.
- Capability inventory: The skill uses file-read operations (
readFile,glob), file-write operations (saveFile), and command-line execution. - Sanitization: The instructions include a requirement to escape health-related text and disable external scripts when generating HTML output.
- [EXTERNAL_DOWNLOADS]: The skill contains a reference to an official FDA resource regarding MRI safety for implants. This is an informational link to a well-known government domain.
Audit Metadata