The agent's log streamer receives the complete process-output snapshot repeatedly while a job runs.
The agent's log streamer receives the complete process-output snapshot repeatedly while a job runs.: a task in MiMo-V2.6-RL-oss: Agentic RL Environments (MiMo RL release). Make this public component safe for that lifecycle: malformed worker settings and repeated startup must be rejected instead of…
The task
Make this public component safe for that lifecycle: malformed worker settings and repeated startup must be rejected instead of creating a streamer that can never drain, and a snapshot that goes backwards must be reported as an error rather than causing a slice panic. For valid…
Part of XiaomiMiMo/MiMo-V2.6-RL-oss.