mailtrap-sending-emails
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFE
Full Analysis
- [DATA_EXFILTRATION]: The skill uses
curlto interact with*.mailtrap.ioAPI endpoints for sending emails. This is consistent with the skill's primary purpose and targets a well-known service. - [EXTERNAL_DOWNLOADS]: The skill references official Mailtrap SDK repositories on GitHub for multiple programming languages. These links point to the service provider's official documentation and source code.
- [INDIRECT_PROMPT_INJECTION]: The skill provides examples for constructing email payloads from application data. This presents a surface where untrusted data could be included in email content.
- Ingestion points: User-provided values for email recipient, subject, and body fields in
SKILL.mdexamples. - Boundary markers: Not present; the examples show direct interpolation into JSON payloads.
- Capability inventory: Network operations via
curland SDKs to send emails externally. - Sanitization: Not explicitly mentioned in the examples, though the skill recommends using official SDKs which often provide structured helpers.
Audit Metadata