Subgraphs (Python)

How UNIQUE_ID and execution behave when your Python node runs inside a subgraph.

UNIQUE_ID in Subgraphs

When a Python node executes inside a subgraph, its UNIQUE_ID becomes a colon-separated execution path instead of a simple number.

Context	UNIQUE_ID	Meaning
Root graph	"789"	Node 789 at top level
One level deep	"45:789"	Node 789 inside subgraph-node 45
Two levels deep	"12:45:789"	Node 789 → subgraph 45 → subgraph 12

def process(self, **kwargs):
    unique_id = kwargs.get('unique_id', '')
    # Always get local ID from the last segment
    local_id = unique_id.split(':')[-1]

    # Use the full unique_id for cache keys and logging
    cache_key = f"my_node_{unique_id}"

Never assume UNIQUE_ID is a simple integer. Always treat it as a string and split on : when you need the local ID.

Execution Flattening

At execution time, the frontend flattens the subgraph hierarchy into a flat list of nodes. Each gets a compound execution ID. The backend sees only the flattened graph — it has no awareness of the subgraph structure.

For runtime subgraph creation in Python, see Node Expansion.

For the frontend developer guide, see Subgraphs.