gamer-com-tw
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill connects to an external MCP server and REST APIs hosted on the vendor's domain (unbrowse.ai) to facilitate interaction with the target website. These connections are used to replay the site's first-party API calls server-side.
- [INDIRECT_PROMPT_INJECTION]: The skill processes content from an external source (gamer.com.tw) which may contain instructions targeting the agent. Ingestion points: Data retrieved from forum boards and news articles via tools like forum_gamer_com_tw__get_b_php and gnn_gamer_com_tw__read_page. Boundary markers: No explicit delimitation or 'ignore instructions' warnings are provided for the ingested web content. Capability inventory: The skill includes tools to perform actions like autofilling logins (unbrowse.browse.act) and potential capabilities for posting or purchasing. Sanitization: The skill instructions do not specify any sanitization or validation logic for the content fetched from the external site.
Audit Metadata