reply-drafting
Pass
Audited by Gen Agent Trust Hub on Sep 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process customer support issues, which are untrusted external inputs that could potentially contain adversarial instructions.
- Ingestion points: Customer issues are read as the primary input for the drafting workflow in
SKILL.md. - Boundary markers: The instructions do not define specific delimiters to isolate the customer input from the agent's instructions.
- Capability inventory: The skill does not include any tools, network access, or shell execution capabilities, which limits the potential impact of an injection.
- Sanitization: There are no explicit input sanitization steps provided.
Audit Metadata