rewrite
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes arbitrary external content to rewrite it into plain language, introducing a risk of indirect prompt injection where malicious instructions embedded in the source text could influence the agent's behavior.
- Ingestion points: Processes user-provided or external content during readability assessment and rewriting steps within
SKILL.md. - Boundary markers: No explicit delimiters, boundary markers, or protective instructions are used to separate user data from system instructions.
- Capability inventory: Invokes internal chains or tools (
readable-contentandlink-text-design) and generates text output, with no direct file-system, network, or command execution capabilities detected. - Sanitization: Lacks input validation, sanitization, or filtering mechanisms to strip instructions out of the text being processed.
Audit Metadata