lg:edge

lg:edge - Edge Management

Manage edges (connections) between nodes in LangGraph graphs including regular edges, conditional edges, and dynamic routing.

Workflow Steps

Add Regular Edge

1. Validate Nodes

   • Check source node exists
   • Check target node exists
   • Verify compatible state schemas

2. Create Edge

   • Add edge to graph definition
   • Update routing logic
   • Add to graph compilation

3. Update Tests

   • Add edge test
   • Update integration tests

4. Update Documentation

   • Update graph diagram
   • Document flow

Add Conditional Edge

1. Validate Configuration

   • Check source node exists
   • Check target nodes exist
   • Validate condition function

2. Generate Condition Function

   • Create routing logic
   • Add condition checks
   • Handle default case

3. Add Conditional Edge

   • Add to graph definition
   • Configure routing map
   • Add error handling

4. Update Tests

   • Test all routing paths
   • Test edge cases
   • Test default routing

5. Update Documentation

   • Document routing logic
   • Add decision tree diagram

Remove Edge

1. Identify Edge

   • Find edge in graph
   • Check for dependencies

2. Remove Edge

   • Remove from graph definition
   • Clean up routing logic
   • Remove condition function (if orphaned)

3. Update Tests

   • Remove edge tests
   • Update integration tests

4. Update Documentation

   • Update graph diagram

Edge Types

Regular Edge

Simple one-way connection.

graph.add_edge("node_a", "node_b")

Flow:

node_a → node_b

Conditional Edge

Branch based on condition.

def route_condition(state: State) -> str:
    """Determine next node based on state."""
    if state["score"] > 0.8:
        return "high_score_path"
    elif state["score"] > 0.5:
        return "medium_score_path"
    else:
        return "low_score_path"

graph.add_conditional_edges(
    "node_a",
    route_condition,
    {
        "high_score_path": "node_b",
        "medium_score_path": "node_c",
        "low_score_path": "node_d"
    }
)