Send Token & Swap Token Agent Benchmark

[dOrgJelli]

Create a multi-agent benchmark that uses the SendToken and SwapToken agents in tandem. The prompts should be sufficiently complex leading to 2+ tasks within the agent's execution plan.

[nerfZael]

We the introduction of autogen (#142) and tests are passing more consistently although the send and swap complex test is at 50%
@dOrgJelli should we close this and open a new one for that case?

[dOrgJelli]

Let's move this work to #166