paw-mkt-dashboard
Warn
Audited by Gen Agent Trust Hub on Apr 7, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill automatically performs shell operations including 'npm install' and 'npm run dev' to initialize and launch generated dashboard servers.- [EXTERNAL_DOWNLOADS]: The skill's generated configuration downloads multiple packages from the public npm registry, including SvelteKit, TailwindCSS, sql.js, and marked.- [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface by recursively ingesting all markdown files within the brand workspace (Ingestion points) into the database and rendering them in the UI. It lacks explicit boundary markers or instruction-ignoring delimiters. The skill has broad capabilities including file system writes and local server execution. While it uses 'marked' for rendering, there is no evidence of sanitization for embedded instructions within the processed markdown files.
Audit Metadata