email-drafting
Pass
Audited by Gen Agent Trust Hub on Apr 10, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill consists exclusively of instructional text and examples for writing emails. It does not include any executable code or scripts.- [NO_CODE]: There are no scripts or tools referenced or included in this skill. It relies entirely on the AI's internal reasoning capabilities.- [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection due to its reliance on user-provided campaign context, but it lacks any exploitable capabilities. 1. Ingestion points: User-supplied campaign context described in Phase 0 of SKILL.md. 2. Boundary markers: Absent for input data. 3. Capability inventory: None; the skill explicitly states it is a pure reasoning skill with no scripts or tools. 4. Sanitization: Absent.
Audit Metadata