email-sequence-design
Pass
Audited by Gen Agent Trust Hub on Sep 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill recommends adding an MCP tool via the command
claude mcp add --transport http manypi https://mcp.manypi.com/mcp -s user. The URLhttps://mcp.manypi.com/mcpis associated with the skill author's domain (manypicom.com). - [INDIRECT_PROMPT_INJECTION]: The skill processes external lead data through template variables such as
{{first_name}}and{{industry}}. This presents a surface for indirect prompt injection if the source data contains malicious instructions. - Ingestion points: Lead data variables in
SKILL.mdused for email template generation. - Boundary markers: None explicitly defined to prevent interpretation of instructions within variable values.
- Capability inventory: The skill references the
ManyPItool for automated email sending and sequence management. - Sanitization: The skill recommends data normalization and the use of fallbacks for missing data, but does not specify sanitization against injection attacks.
Audit Metadata