Orchestrate
Category: Agent
Node id: orchestrate
Provides workflow orchestration with LLM-powered planning, task decomposition, routing, and synthesis
What it does
- LLM-powered execution plan generation with ordered steps
- Complex task decomposition into actionable sub-tasks
- Intelligent task routing to available processing nodes
- Multi-result synthesis with conflict resolution
- Dependency and constraint-aware planning
Use cases
- Decomposing complex clinical workflows into executable steps
- Routing tasks to the most appropriate pipeline nodes
- Synthesizing results from parallel sub-task execution
- Generating execution plans for multi-stage data processing
Runtime
- Performance tier:
variable
Input ports
| Name | Type | Required | Description |
|---|---|---|---|
objective | text | Yes | The objective or task to orchestrate |
context | text | No | Additional context for planning |
constraints | text | No | Constraints for plan generation |
sub_results | json | No | Sub-task results for synthesis operation |
Output ports
| Name | Type | Required | Description |
|---|---|---|---|
plan | json | No | Generated execution plan (plan operation) |
tasks | json | No | Decomposed sub-tasks (decompose operation) |
routing | json | No | Task routing decision (route operation) |
synthesis | text | No | Synthesized result (synthesize operation) |
operation | text | No | Operation that was performed |
objective | text | No | Original objective |
Settings
Basic
| Setting | Name | Type | Required | Default | Allowed values | Description |
|---|---|---|---|---|---|---|
| LLM Endpoint | llm_endpoint | string | Yes | Base URL of the LLM API | ||
| Model | model | string | No | gpt-4o | Model name to use | |
| Operation | operation | enum | No | plan | plan, decompose, route, synthesize | Orchestration operation to perform |
Auth
| Setting | Name | Type | Required | Default | Allowed values | Description |
|---|---|---|---|---|---|---|
| API Key | api_key | secret | No | API key for the LLM provider |
Advanced
| Setting | Name | Type | Required | Default | Allowed values | Description |
|---|---|---|---|---|---|---|
| Available Nodes | available_nodes | json | No | Available processing nodes for routing operation |
Errors
| Code | Description |
|---|---|
ENDPOINT_REQUIRED | llm_endpoint is not configured |
OBJECTIVE_REQUIRED | No objective input was provided |
UNSUPPORTED_OPERATION | Operation must be plan, decompose, route, or synthesize |
LLM_ERROR | LLM returned an error during orchestration |
Runtime notes
- Output keys vary by operation: plan returns {plan, operation, objective}; decompose returns {tasks, ...}; route returns {routing, ...}; synthesize returns {synthesis, ...}
- plan and tasks outputs are JSON-parsed from the LLM response; if parsing fails, the raw text is returned
- The synthesize operation expects sub_results input as an array of prior task outputs
- The route operation requires available_nodes config to list candidate nodes for routing decisions
See Build any workflow for how to put nodes together and Ports, types and data flow for how port types connect.
Etwas unklar? Senden Sie Feedback zur Dokumentation per E-Mail..