audit-responsive
Pass
Audited by Gen Agent Trust Hub on Oct 24, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill analyzes external web application content and local source code, which serves as a potential vector for indirect prompt injection. A malicious application or code snippet could attempt to influence the agent's behavior during the audit or implementation phases.
- Ingestion points:
SKILL.mdPhase 0 and Phase 1 usergto read local files and Playwright to capture snapshots and screenshots of the application at<app-url>. - Boundary markers: The instructions do not define specific delimiters to separate audited data from agent instructions, which may lead to the agent following instructions embedded in the UI or code.
- Capability inventory: The agent has capabilities for file reading (
rg), browser automation (playwright), and file modification (Phase 4 implementation). - Sanitization: No sanitization logic is present for the data ingested during the audit phase.
- [EXTERNAL_DOWNLOADS]: The skill uses
npxto download and execute the@playwright/clipackage from the official npm registry to perform browser automation tasks.
Audit Metadata