experiment-queue
Pass
Audited by Gen Agent Trust Hub on May 18, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill constructs and executes shell commands on remote SSH hosts to manage experiment lifecycles. This includes using
nohupandscreento ensure job persistence across sessions, which is standard behavior for the described use case. - [PROMPT_INJECTION]: The skill generates experiment manifests by parsing user-provided natural language and YAML grid specifications. This processing of untrusted input creates a surface for indirect prompt injection, where malicious instructions could be embedded in experiment parameters (e.g., in grid field values). However, the capability is restricted to the context of shell command construction for experiment scripts.
Audit Metadata