The Agent Skills Directory

[Unverifiable Dependencies & Remote Code Execution] (LOW): The skill documentation states that Playwright will be installed automatically. This process involves fetching code from the npm registry, which introduces a minor supply chain risk when versions are not pinned or verified.
[Prompt Injection] (LOW): The skill is susceptible to indirect prompt injection because it processes untrusted data from web pages and browser logs.
Ingestion points: Console logs captured in test-helper.js and page content/titles accessed in SKILL.md examples.
Boundary markers: Absent. The skill does not use delimiters to distinguish between system instructions and untrusted browser data.
Capability inventory: File system writing via captureScreenshot and network access via page.goto and browser automation.
Sanitization: Absent. Data retrieved from the browser is logged and stored without being sanitized or escaped.
[Data Exposure & Exfiltration] (LOW): The skill captures and saves browser screenshots and logs to the local file system. While this is a core functionality for debugging, it may inadvertently record sensitive information or credentials displayed in the application being tested.

webapp-testing