// Wraps user steering input with context signals for the model during a wild loop session
| name | wild_v2_steer |
| description | Wraps user steering input with context signals for the model during a wild loop session |
The user is actively steering this autonomous session. They have paused the loop to provide the following guidance. You MUST prioritize this input over your current plan.
{{goal}}
tasks.md to reflect any changes the user requests. If they want you to stop a particular approach, mark those tasks as cancelled.The user trusts you but wants to course-correct. Take their input seriously and adapt accordingly.
System prompt for handling experiment alerts. Provides diagnosis guidance, GPU wrapper context, action suggestions, and structured response from allowed choices.
Default system prompt for agent chat mode. Provides identity, environment context, compute awareness, API-driven job submission, and workflow reflection.
Generates a structured experiment plan with compute-aware recommendations and saves it via the plan API endpoint.
Ports new models into FastVideo with strict numerical alignment to official implementations. Use when adding a FastVideo model/pipeline, porting an official or Diffusers checkpoint, or debugging parity/alignment.
Single source of truth protocol for Wild V2 preflight, sweep/run auditability, GPU discovery, and parallel scheduling
Protocol for GPU discovery and parallel run scheduling across local GPU and Slurm clusters