prompt-optimizer
Pass
Audited by Gen Agent Trust Hub on Mar 26, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: Comprehensive analysis of the skill reveals no instructions for data exfiltration, remote code execution, or unauthorized persistence. The skill's role is strictly advisory, as explicitly stated in its instructions.
- [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface as it ingests untrusted user draft prompts for analysis. However, the risk is negligible due to the absence of execution-oriented tools or network capabilities.
- Ingestion points: User-drafted prompts provided for optimization (SKILL.md).
- Boundary markers: The skill does not employ delimiters to wrap the untrusted user content.
- Capability inventory: Capabilities are restricted to text analysis and reasoning; no file-writing, network-accessing, or shell-executing commands are present.
- Sanitization: No validation or sanitization of the input prompt is performed.
Audit Metadata