quant-buddy-view
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the subprocess module to manage its internal workflows. This involves executing localized Python scripts and Node.js utilities for tasks such as bridging with the peer quant-buddy-skill or initiating the verify_page.mjs suite. These calls are scoped to internal components and do not execute arbitrary user-supplied strings.
- [DYNAMIC_EXECUTION]: Browser-based automation via Playwright is used in several scripts (e.g., capture_rendered_html.mjs) to conduct smoke tests and capture snapshots of generated dashboards. This ensures published pages are visually correct. The skill also employs lazy-loading for internal modules using importlib, which is used for standard architectural organization.
- [EXTERNAL_DOWNLOADS]: The skill includes a self-update mechanism that fetches releases from its official repository on GitHub, an established service. It also utilizes legitimate public CDNs to load visualization libraries like ECharts. These operations are performed using standard secure transport protocols.
- [INDIRECT_PROMPT_INJECTION]: The skill converts external files (such as HTML or PDF documents) into interactive dashboards. While this creates a potential surface for indirect prompt injection via the content of those files, the skill implements data redaction and task-level isolation to mitigate risks. Furthermore, its snapshot capture logic specifically blocks the browser from reading arbitrary local files beyond the target document.
Audit Metadata