customer-journey-mapper
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data in the form of customer profiles and product descriptions, which defines a potential surface for indirect prompt injection if the inputs contain malicious instructions. However, the workflow directs the agent to treat this as data for map generation. Ingestion points are defined in the 'Input Contract' section of SKILL.md. The capability inventory includes generating Markdown tables and visualization scripts. No specific boundary markers or sanitization steps are defined in the instructions to mitigate embedded instructions within the processed data.
- [DYNAMIC_EXECUTION]: The skill provides instructions for generating visualization scripts using external libraries to render emotion curves. Evidence includes references to using matplotlib for PNG generation and Chart.js for web-based rendering in the references/emotion-rendering.md file. This is a legitimate use case for the skill's primary function of data visualization and is not indicative of malicious intent.
Audit Metadata