linux-server-ssh
Fail
Audited by Gen Agent Trust Hub on Jun 13, 2026
Risk Level: HIGHREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill recommends installing the 'box' CLI using a piped shell command: 'curl -fsSL https://box.ascii.dev/install | sh'. This method executes a remote script without any verification of its content or integrity.
- [EXTERNAL_DOWNLOADS]: The skill depends on tools and assets hosted at 'box.ascii.dev', a domain that is not identified as a trusted vendor or well-known technology service provider.
- [COMMAND_EXECUTION]: The skill provides the ability to execute arbitrary commands on remote servers through 'box ssh', granting the agent significant control over external infrastructure.
- [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface when processing remote server output.
- Ingestion points: Data returned from the remote server via 'box ssh' and 'box info' in SKILL.md.
- Boundary markers: There are no delimiters or instructions provided to help the agent distinguish between its instructions and the output from the remote host.
- Capability inventory: The skill can create instances and execute arbitrary code on remote infrastructure.
- Sanitization: No sanitization or validation of the server's output is performed before it is integrated into the agent's context.
Recommendations
- HIGH: Downloads and executes remote code from: https://box.ascii.dev/install - DO NOT USE without thorough review
Audit Metadata