wordpress-full-compromise
Pass
Audited by Gen Agent Trust Hub on Jul 31, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill uses
curlandpython3subprocess calls withinreferences/mailinator-reset-workflow.mdto interact with target systems and process data. Thecurlcommands utilize the-kflag, which bypasses SSL certificate verification. - [DATA_EXFILTRATION]: The skill performs network operations to
www.mailinator.comto retrieve email content for registration testing. This interaction with a well-known service is consistent with the skill's primary purpose. - [PROMPT_INJECTION]: The skill contains an indirect prompt injection surface. 1. Ingestion points: Data is fetched from external email bodies via the Mailinator API in
references/mailinator-reset-workflow.md. 2. Boundary markers: The instructions do not provide delimiters to separate retrieved email content from agent instructions. 3. Capability inventory: The agent has capabilities for network operations (curl) and data processing (python3). 4. Sanitization: The skill extracts specific keys using regular expressions but does not perform comprehensive sanitization of the external content.
Audit Metadata