copy-editing
Pass
Audited by Gen Agent Trust Hub on Nov 1, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill consists entirely of markdown instructions defining a framework for copy editing. It does not include any executable scripts, tool configurations, or command-line operations.- [NO_CODE]: Analysis of the skill file confirms the absence of any code (Python, JavaScript, shell scripts) or configuration files that could execute logic on a host system.- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process user-supplied marketing copy, creating a surface for indirect prompt injection. However, the risk is negligible as the skill possesses no capabilities for exfiltration, file modification, or system interaction. Evidence chain: 1. Ingestion points: User-provided text in SKILL.md. 2. Boundary markers: Absent. 3. Capability inventory: None. 4. Sanitization: Absent.
Audit Metadata