business-card
Warn
Audited by Gen Agent Trust Hub on Mar 31, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill explicitly instructs the agent to use shell commands, including the
sedutility for bulk editing of SVG files and theqrencodetool for creating QR codes. - [DATA_EXFILTRATION]: The skill workflow requires collecting sensitive Personally Identifiable Information (PII) such as full names, home/office addresses, and personal phone numbers, which are stored in the user's local project files.
- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection as it takes user-provided contact details and embeds them directly into SVG and HTML files without instructions for sanitization or output encoding.
- Ingestion points: Phase 0 Interview in
SKILL.mdwhere the user provides contact details. - Boundary markers: None present to delimit user input from the file structure.
- Capability inventory: Shell command execution via
sedandqrencode, and the ability to write to SVG and HTML files. - Sanitization: No sanitization or validation logic is defined for the incoming data.
Audit Metadata