skills/elastic/kibana/kbn-ui-package/Gen Agent Trust Hub

kbn-ui-package

Warn

Audited by Gen Agent Trust Hub on Aug 28, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill performs bulk search-and-replace operations across the entire repository (src/, x-pack/, packages/) using sed -i based on user-influenced variables (oldName, packageName). These operations are high-impact and affect every TypeScript file matching the search criteria.
  • [COMMAND_EXECUTION]: Shell commands utilize user-provided inputs like folderName and description without explicit sanitization, creating a surface for command or logic injection in operations like git mv and sed.
  • [INDIRECT_PROMPT_INJECTION]: The skill consumes data from existing repository files (package.json, index.ts) to influence its behavior and code generation. Ingestion points include local project files from a user-specified path, and the instructions lack boundary markers or sanitization for values extracted from these files before they are interpolated into templates or shell commands.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Aug 28, 2026, 09:49 PM
Security Audit — agent-trust-hub — kbn-ui-package