Swarm

Peer agents pass control via a handoff tool. OpenAI Swarm reference (late 2024, experimental). Anthropic Agent Teams (Feb 2026) is the production answer that improved on the original swarm idea by adding lightweight coordination.

   prompt ──► agent A
                 │
                 │ handoff(B, payload)
                 ▼
              agent B
                 │
                 │ transfer_to_C(typed_args)
                 ▼
              agent C ──► final output
                 ▲
                 │ cycle detection: A→B→A→B kills the loop
                 │ max_handoffs caps total depth

Pattern

1. Setup. N peer Agent instances; one designated as the entry agent (receives the first user message).
2. Active turn. The active agent runs to completion with one or more handoff tools injected. The model can call them (or not) freely during the turn.
3. Detect handoff. After the agent’s turn ends, Swarm checks whether a handoff tool was called. If yes, switch active agent to the named target.
4. Loop or terminate. Continue handoffs until either: (a) an agent finishes without calling a handoff tool, (b) max_handoffs reached, or (c) cycle detected.

Two handoff styles

from loomflow import Agent
from loomflow.architecture import Handoff
from loomflow.team import Team
 
triage  = Agent("Triage incoming requests.",  model="claude-opus-4-7")
billing = Agent("Handle billing.",            model="claude-opus-4-7", tools=[...])
tech    = Agent("Handle technical issues.",   model="claude-opus-4-7", tools=[...])
 
# Style 1: a single generic ``handoff(name, payload)`` tool
team = Team.swarm(
    agents={"triage": triage, "billing": billing, "tech": tech},
    entry_agent="triage",
    model="claude-opus-4-7",
)
 
# Style 2: typed handoffs — wrap each peer in ``Handoff`` so you get
# per-target ``transfer_to_<name>`` tools with structured arguments
team = Team.swarm(
    agents={
        "triage": triage,
        "billing": Handoff(billing, input_type=BillingHandoff),
        "tech": Handoff(tech, input_type=TechHandoff),
    },
    entry_agent="triage",
    model="claude-opus-4-7",
)

input_type= is a Pydantic model. The handoff tool’s schema is derived from it, so the model emits structured payloads instead of free-form strings.

Cycle detection

Swarm tracks the handoff sequence per run. A repeating chain (A→B→A→B) trips cycle detection and terminates the run with interrupted=True, interruption_reason="swarm_cycle". Disable with cycle_detection=False if you genuinely want long handoff sequences.

When (sparingly) to use Swarm

• Exploratory pipelines where the flow shape will change weekly.
• Research-mode systems where you’re testing what handoff patterns emerge.

For production multi-agent: use Supervisor or Router.