enrichment-design
Pass
Audited by Gen Agent Trust Hub on Mar 3, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: A thorough review of the skill logic and reference files found no indicators of malicious activity. The skill operates within its stated purpose of assisting with enrichment column design.
- [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface as it processes external hypothesis sets and company lists to generate prompts. 1. Ingestion points: Hypothesis sets and target lists read by
SKILL.md. 2. Boundary markers: No explicit delimiters or boundary markers are included in the generated prompt templates. 3. Capability inventory: Generates research instructions for agents (research_pro,llm) that have web research and reasoning capabilities. 4. Sanitization: The skill does not perform validation or escaping of the{input}variable or external hypotheses before interpolation into prompts.
Audit Metadata