ui-theme-designer-help
Pass
Audited by Gen Agent Trust Hub on Jul 29, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill fetches deliverable metadata and topic content from https://help.sap.com. This is the official documentation portal for SAP, the skill's author, and is a trusted source for the intended functionality.
- [COMMAND_EXECUTION]: Procedural steps use curl for network requests and jq, sed, and tr for processing JSON and formatting text. These commands are transparently defined and restricted to formatting the documentation for the agent's response.
- [PROMPT_INJECTION]: The skill processes data from external web pages, creating a surface for indirect prompt injection.
- Ingestion points: Documentation content is retrieved from help.sap.com in SKILL.md.
- Boundary markers: The instructions do not define specific delimiters for the retrieved content.
- Capability inventory: Capabilities are limited to curl, jq, sed, and tr. There are no file system write or code execution permissions.
- Sanitization: HTML tags are stripped using sed before the content is processed by the agent.
Audit Metadata