Detect Circular Money Flows in a Payment Transaction Graph

Problem

Given a list of payment transactions [senderId, receiverId, amount], detect all cycles in the transaction graph — these indicate circular money flows that may signal fraud or settlements.

Example

transactions = [["A","B",100],["B","C",200],["C","A",150],["D","E",50]]
Output: [["A","B","C"]]  -- cycle A→B→C→A

Constraints

• Up to 10^5 transactions
• Account IDs are strings (UUIDs in production)

Juspay-specific context

In payment systems, circular flows can represent valid settlement rings (netted clearing) or fraud chains. Your solution should return the full cycle path, not just a boolean.

Expected approach

DFS with three-color marking (unvisited/in-stack/visited). Track the path to reconstruct cycles.