postmark-automation
Pass
Audited by Gen Agent Trust Hub on Mar 5, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADSNO_CODE
Full Analysis
- [PROMPT_INJECTION]: The skill defines a potential surface for indirect prompt injection (Category 8) by ingesting and processing external data.
- Ingestion points: Processes email templates via
POSTMARK_GET_TEMPLATEand delivery feedback/bounce data viaPOSTMARK_GET_BOUNCESandPOSTMARK_GET_SPAM_COMPLAINTS. - Boundary markers: No specific delimiters or instructions to ignore embedded commands are defined in the orchestration logic.
- Capability inventory: High-impact capabilities include sending batch emails via
POSTMARK_SEND_BATCH_WITH_TEMPLATES, editing templates viaPOSTMARK_EDIT_TEMPLATE, and modifying server settings viaPOSTMARK_EDIT_SERVER. - Sanitization: No explicit sanitization or validation logic is provided for template variables or external inputs before they are used in tool calls.
- [EXTERNAL_DOWNLOADS]: The skill requires a connection to an external MCP server at
https://rube.app/mcp. This is a standard integration pattern for using Rube-based toolkits and points to a known service provider. - [NO_CODE]: This is a documentation-only skill that guides an agent on how to use external tools. It contains no local scripts, binaries, or executable code, which significantly reduces the local attack surface.
Audit Metadata