quality-rubric
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns, prompt injections, or dangerous execution capabilities were found within the instructions or the reference markdown files.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted draft content. Evidence chain: 1. Ingestion points: draft and report files in SKILL.md. 2. Boundary markers: Absent. 3. Capability inventory: Limited to text-based report generation and routing recommendations. 4. Sanitization: Absent. The lack of dangerous capabilities like shell execution or network access makes this surface safe.
Audit Metadata