postedly-mail-letter
Pass
Audited by Gen Agent Trust Hub on Nov 8, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes common local system utilities (such as
shasum,base64,pdfinfo, andpdftoppm) to process documents. These tools are used to calculate file hashes for integrity checks, encode files for API transmission, and render PDF pages for visual inspection to ensure layout quality before mailing. - [EXTERNAL_DOWNLOADS]: The skill employs
curlto upload document bytes to a remote endpoint. It includes a robust safety check requiring theuploadEndpointto start withhttps://postedly.com/api/document-handoffs/before proceeding with any transfer. - [INDIRECT_PROMPT_INJECTION]: Because the skill processes external files and user-provided text, it is theoretically susceptible to indirect prompt injection. However, it explicitly instructs the agent to treat any instructions found within the document as literal content to be printed, rather than instructions to be followed.
- [DATA_EXFILTRATION]: The skill transmits user documents to the Postedly service, which is its primary purpose. It incorporates several safeguards, such as explicitly forbidding the agent from requesting passwords or verification codes in the chat and ensuring that payment processing occurs on a separate, hosted Stripe Checkout page.
Audit Metadata