Skip to content

Self-improving AI Employees for everyone

3.1 KB
Raw Blame

Orchestrator Agent (orchestrator_agent)

Field Value
Category specialized_agents
Plugin server/nodes/agent/orchestrator_agent/__init__.py -> _specialized.py::SpecializedAgentBase.execute_op (dispatch via BaseNode.execute())
Team lead yes -- input-teammates handle enabled (team_lead_agent_handles())
Tests server/tests/nodes/test_specialized_agents.py

Purpose

Team-lead agent that coordinates agents wired through input-teammates. Task Manager is intrinsically bound and is the only model-facing delegation interface. Internal delegate descriptors resolve authorized assignees after a durable assign_task record exists.

What is unique to this node

  • input-teammates handle (extra input beyond the shared 5; declared via team_lead_agent_handles() in _handles.py).
  • No extra parameters: uses the same SpecializedAgentParams as every other agent. There is no teamMode / maxConcurrent field — those were removed (or never existed on the current Pydantic model).
  • tool_description declared so a parent team lead delegating to this node gets a meaningful tool string ("ONE-SHOT delegation ...").
  • Team lead detection: prepare_agent_call (in _inline.py) checks node_type in TEAM_LEAD_TYPES = {'orchestrator_agent', 'ai_employee'} and calls collect_teammate_connections to build internal teammate descriptors.
  • Task control panel: selecting the node opens its non-removable Task Manager middle panel.

Teammate collection

collect_teammate_connections(node_id, context, database) in server/services/plugin/edge_walker.py:

  1. Scans context.edges for edge.target == node_id and edge.targetHandle == 'input-teammates'.
  2. Resolves the source node in context.nodes.
  3. Filters to node_type in AI_AGENT_TYPES.
  4. Loads database.get_node_parameters(source_id) for each teammate.
  5. Returns a list of {node_id, node_type, label, parameters} dicts.

prepare_agent_call walks each teammate's own input-tools edges to populate capability descriptions. The LLM receives connected node IDs and calls task_manager(operation="assign_task", ...); direct delegate tools are hidden.

Behaviour

Inputs, parameters, outputs, logic flow -- see Generic Specialized Agent Pattern. The only difference is the teammate expansion above.

Edge cases

  • Non-AI_AGENT_TYPES nodes wired to input-teammates are silently skipped.
  • Concurrency is root-scoped and defaults to three active descendants. Additional durable assignments remain queued.
  • When zero teammates are connected, the node behaves identically to any other specialized agent.

Related