infrastructure-lambda-labs
Pass
Audited by Gen Agent Trust Hub on Nov 3, 2026
Risk Level: SAFECOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes a wide range of shell commands for system diagnostics, package management, and infrastructure control both locally and on remote GPU instances.
- [EXTERNAL_DOWNLOADS]: The skill downloads official installation scripts and software packages from trusted and well-known sources, such as Lambda Labs and Anaconda.
- [REMOTE_CODE_EXECUTION]: Includes utilities to execute arbitrary commands on remote servers via SSH and documents the use of piped shell scripts for official stack installation from the service provider's domain.
- [INDIRECT_PROMPT_INJECTION]:
- Ingestion points: Remote training logs are read from instances via the monitor_job function in references/advanced-usage.md.
- Boundary markers: There are no explicit delimiters or warnings used to prevent the agent from following instructions potentially embedded in log files.
- Capability inventory: The skill has the ability to execute local shell commands, perform remote command execution, and interact with the Lambda Labs API to manage cloud resources.
- Sanitization: The scripts do not perform sanitization or validation of the log data being monitored or the commands being constructed for remote execution.
Audit Metadata