box
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill directs users to install the
@membranehq/clipackage globally via NPM. While this is a tool provided by the platform vendor, installing global packages involves executing code from an external registry. - [COMMAND_EXECUTION]: The skill uses shell commands through the
membraneCLI to perform authentication, connection management, and Box API actions. This is a core part of its functionality. - [INDIRECT_PROMPT_INJECTION]: The skill processes data from Box, such as file versions, folder collaborations, and group information, which could contain malicious instructions designed to influence the agent's behavior.
- Ingestion points: Data fetched from Box via
membrane action runandmembrane requestcommands as described in SKILL.md. - Boundary markers: No specific delimiters or warnings to ignore instructions within the retrieved data are implemented.
- Capability inventory: The skill allows shell command execution and network requests via the
membraneCLI (SKILL.md). - Sanitization: There is no evidence of sanitization or filtering of the content retrieved from the Box API.
Audit Metadata