apple
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The README.md and SKILL.md files instruct the user to install the skill using
npx skills add https://unbrowse.ai, which downloads and executes code from the vendor's domain to set up the environment. - [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external web pages on apple.com, creating a surface for potential injection attacks. 1. Ingestion points: Data enters via the
apple_com__render_pageandapple_com__read_pagetools described in SKILL.md. 2. Boundary markers: There are no explicit delimiters or instructions provided to the agent to treat the ingested web content as untrusted data. 3. Capability inventory: The skill has the capability to perform further network requests based on link discovery (crawling). 4. Sanitization: No specific sanitization or filtering logic for the ingested markdown/text is present in the skill instructions. - [DATA_EXFILTRATION]: The skill is configured to use a remote MCP server at
https://unbrowse.ai/mcp/apple.com. All user queries and browsing paths are sent to this third-party infrastructure to be proxied to the target site.
Audit Metadata