skills/vm0-ai/vm0-skills/zeptomail/Gen Agent Trust Hub

zeptomail

Pass

Audited by Gen Agent Trust Hub on Apr 17, 2026

Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes system commands including curl to interact with the ZeptoMail API and base64 to process file attachments. These operations are aligned with the stated purpose of sending emails.
  • [DATA_EXFILTRATION]: The skill facilitates the transmission of data and local files to the ZeptoMail API. While interacting with a well-known service provider, this functionality establishes a channel for external data transmission. The skill also suggests writing request payloads to /tmp/zeptomail_request.json, which is a common practice for handling CLI data but could potentially expose data to other local processes.
  • [EXTERNAL_DOWNLOADS]: The skill makes network requests to well-known Zoho ZeptoMail API endpoints to provide its core email delivery services.
  • [PROMPT_INJECTION]: The skill processes user-provided content for email components such as subjects, bodies, and merge fields. This represents an indirect prompt injection surface because it ingests untrusted data into the agent's context without specific boundary markers or sanitization, though the risk is confined to the email generation and delivery workflow.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 17, 2026, 04:44 PM