Agent and model settings

Configure a scheduled prompt with repeat, timer, or date-based execution.

Search shared skills and edit their instructions beside the file list.

Keep the agent task plan visible with active work and completed items.

Review completed scheduled runs, inspect their outputs, and archive results.

Monitor concurrent background commands and open each job’s output.

Inspect context usage and configure which status details stay visible.

Browse MCP configurations and edit the server connection definition.

Manage provider accounts, session limits, weekly usage, and switching.

Browse recurring jobs alongside their prompts, schedules, and recent runs.

See each child agent’s model, task, token usage, and elapsed time.

Visual agent tree with editable branch steps, model selection, a docked inspector, pan and zoom, automatic layout, test runs and JSON export.

Visual agent builder with trigger, instruction, model, knowledge, tool, guardrail, and handoff nodes plus test, version, and publish controls.