prompt-engineering
Pass
Audited by Gen Agent Trust Hub on Sep 6, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze and process user-provided prompts and LLM templates, which are untrusted data sources.
- Ingestion points:
SKILL.md(processing user-supplied prompts for auditing or porting). - Boundary markers: The skill includes explicit instructions to use XML delimiters (e.g.,
<input>,<context>) to fence off variable input from instructions. - Capability inventory: None. The skill consists entirely of instructional markdown and does not utilize any tools for network access, file system modification, or command execution beyond fetching official documentation.
- Sanitization: Recommends structural delimitation to prevent instruction confusion.
- [EXTERNAL_DOWNLOADS]: The skill references and suggests fetching live documentation from official vendor platforms including Anthropic, OpenAI, and Google. These are well-known, trusted sources for model guidance.
Audit Metadata