remote-compute-ssh
Pass
Audited by Gen Agent Trust Hub on Sep 8, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill is designed to facilitate remote host setup and research job submission through SSH and Slurm. This allows the agent to execute shell commands and scripts on external laboratory or cluster systems as its primary function.- [INDIRECT_PROMPT_INJECTION]: The skill harvests data from remote job logs and status updates, creating a vulnerability to untrusted content.
- Ingestion points: The workflow involves reading remote stdout, stderr, and artifact checksums back into the Feynman workspace (SKILL.md).
- Boundary markers: There are no explicit instructions to use delimiters or ignore embedded instructions when processing harvested remote content.
- Capability inventory: The skill includes job submission, log harvesting, and remote environment configuration capabilities.
- Sanitization: The instructions do not define any sanitization or validation protocols for the harvested remote job output.
Audit Metadata