Agentic Settings
The Agentic Settings page configures shared defaults for specialist agent runtime behavior.
This is where operators shape the general execution posture for specialists without rebuilding an image or editing every workflow by hand.
What Lives Here
Section titled “What Lives Here”- specialist agent container defaults
- workload limits and backlog posture
- specialist and orchestrator context retention
- loop safeguards and retry ceilings
- runtime-side timeouts and execution defaults
Why It Matters
Section titled “Why It Matters”These settings define the shared baseline for how specialist agents behave once the platform has routed work. They are part of the product’s operating posture, not just internal implementation knobs.
Important Boundary
Section titled “Important Boundary”Execution environment records live on Platform -> Environments. Agentic Settings controls runtime defaults and loop posture, not the task-image catalog itself.
That separation is deliberate. One page answers “where does this work run?” The other answers “how should specialists generally behave once they are running?”
How It Connects To The Rest Of The System
Section titled “How It Connects To The Rest Of The System”Agentic Settings is the shared runtime baseline that sits underneath the more specific assignment pages:
- Environments defines the execution profile or image a specialist will run in
- Specialists, Skills, And Models defines the role-level contract for a specific specialist
- Platform Settings defines the control-plane timing and supervision around that execution
- Runtime Operations is where the whole runtime-facing posture makes sense as one operating model
This page answers “what should specialist execution generally look like?”, while the other pages answer “who runs?”, “where do they run?”, and “how does the platform supervise them?”