pdforge
Pass
Audited by Gen Agent Trust Hub on Apr 17, 2026
Risk Level: SAFEDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [DATA_EXFILTRATION]: The skill performs network requests to api.pdfnoodle.com and api.pdforge.com using curl. These operations transmit user-provided data or HTML content to the external PDF generation service, which is the intended primary function of the skill.
- [PROMPT_INJECTION]: The skill has a surface for indirect prompt injection because it ingests and processes untrusted HTML and JSON data to send to the PDF generation API.
- Ingestion points: External HTML strings and data objects defined in the request examples.
- Boundary markers: There are no explicit delimiters or instructions provided to the agent to treat user data as untrusted content.
- Capability inventory: The skill utilizes file writing to /tmp/ and network communication via curl.
- Sanitization: No sanitization or validation of the input HTML or template data is described in the documentation.
Audit Metadata