box
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses
npx @useclawlink/clito interact with the ClawLink service and perform actions in Box, such as connecting accounts and running file management commands. - [EXTERNAL_DOWNLOADS]: The skill utilizes the
@useclawlink/clipackage, which is fetched from the NPM registry at runtime when invoked vianpx. - [INDIRECT_PROMPT_INJECTION]: The skill processes content and metadata from external Box storage, which constitutes a vulnerability surface for instructions embedded in files or folder names.
- Ingestion points: Data retrieved from Box via actions like reading file content or listing folder structures.
- Boundary markers: Absent; the skill does not define specific delimiters to isolate external data from the agent's instructions.
- Capability inventory: The skill provides significant capabilities, including file modification, user/group management, and sharing link creation within Box.
- Sanitization: No specific sanitization or filtering of external content is mentioned in the skill instructions.
Audit Metadata