premium-commercial-writing
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external data, including client briefs, source packs, and existing drafts for rewriting. This presents a surface for indirect prompt injection, where an attacker could embed instructions within the processed data to influence the AI's output or behavior.
- Ingestion points: Client-provided briefs, source packs, and drafts referenced in the Workflow and Required Inputs sections of SKILL.md.
- Boundary markers: The skill contains instructions to flag gaps in proof and obtain explicit client authority for sensitive actions, which acts as a procedural boundary.
- Capability inventory: The skill's scope is restricted to reading and searching only; it lacks built-in capabilities for file-writing or direct network exfiltration.
- Sanitization: There are no technical filters or sanitization steps mentioned for the ingested text.
- [SAFE]: The skill references an external resource for design constraints located at
github.com/peterbamuhigire/chwezi-design-engine. This URL originates from the vendor's repository and is used for legitimate documentation purposes.
Audit Metadata