prompt-engineering
Pass
Audited by Gen Agent Trust Hub on Apr 7, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill includes instructions for the agent to check for the presence of recommended companion skills by listing contents of specific local directories (e.g.,
~/.claude/skills/). This command execution is a benign helper feature used to offer relevant installation commands for missing tools. - [SAFE]: The skill handles untrusted data ingestion via templates for RAG and few-shot examples. It follows security best practices by explicitly instructing the agent to use structural delimiters (like XML tags or triple dashes) and schema validation to prevent indirect prompt injection and hallucinations.
Audit Metadata