What problem does it solve? Long-running screening and confirmation experiments run as background jobs, and checking them manually risks corrupting the research pipeline by acting on partial logs or misreading process handles. This Skill provides a safe, read-only way to check job status and detect completion without changing any scientific state. ## Core Features & Use Cases - Status Inspection: Resolves the registered DSH background job id and reports whether the job is running, completed, failed, or missing, along with the last meaningful metric or error. - State Safety Guarantees: Never creates goal rounds, increments attempts, rejects candidates, changes phases, or triggers reviewers based on partial logs. - Completion Handoff: When a job finishes, passes raw artifact paths to experiment-run for metric collection and routing; on infrastructure failure, preserves the candidate and plan while the launch issue is fixed. - Use Case: A screening job wakes your session via a job event. Use this Skill to confirm it completed, retrieve the artifact paths, and hand them off for metric collection instead of polling logs and guessing at results. ## Quick Start Check the status of the currently registered background experiment job and tell me whether it completed, failed, or is still running.
