The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes untrusted external data by navigating to URLs and analyzing their content. This creates an attack surface for indirect prompt injection, where a malicious website could embed instructions to influence the agent's next steps.
Ingestion points: Processes external content via the url parameter in page.goto(url) calls.
Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present in the documentation.
Capability inventory: Navigates to external websites using Playwright and writes screenshot files to the local screenshots/ directory.
Sanitization: No sanitization or validation of the fetched HTML/metadata is described in the prompt logic.
[COMMAND_EXECUTION]: The skill references a local Python script scripts/capture_screenshot.py for automated capture tasks. This indicates the skill relies on executing local executable code to perform its primary function.

seo-visual