antislop-copywriting
Pass
Audited by Gen Agent Trust Hub on Aug 31, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and edit user-supplied prose (marketing copy, landing pages, etc.). This creates an inherent surface for indirect prompt injection where malicious instructions could be embedded in the data being processed to influence the agent's behavior during the editing task.
- Ingestion points: User-provided prose and copy projects referenced in SKILL.md.
- Boundary markers: The skill references an external core system (antislop.md) and a 'Delivery Gate' process for verification, though no explicit delimiter-based sanitization is described in this specific file.
- Capability inventory: The skill uses Read, Write, Edit, Glob, and Grep tools to manage and modify files.
- Sanitization: The skill explicitly instructs the agent to 'Never invent facts' (R-17, R-36, R-38) and requires evidence for claims, which mitigates the risk of data fabrication or hallucination during processing.
Audit Metadata