email-write
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from project-level files (
email-profile.mdandreferences/copy-frameworks.md) to generate email templates, which introduces a surface for indirect instructions to influence the agent. - Ingestion points: Data is read from the local file system during the 'Gather Context' and 'Select Copy Framework' steps in
SKILL.md. - Boundary markers: The skill uses Markdown code blocks for output but lacks specific instructions to disregard potentially malicious commands embedded in the source files or user-provided personalization data.
- Capability inventory: The skill uses the
Read,Write,Grep, andGlobtools to interact with local files. - Sanitization: There are no defined procedures for escaping or validating external input before it is interpolated into the final HTML or plain-text email bodies.
Audit Metadata