I named the board BaDiTo after Battery Discharge tool. The first version of the board basically consist of power resistors, a mosfet and a comparator and a LED. Using one rocker switch you can choose between 1.5A load (2 resistors in parallel) or 3A load (4 resistors in parallel) which equals about 0.5C and 1C discharge current. The other rocker switch is to choose between two discharge voltages, 3.5V or 3.0V. The LED is on during discharging and it goes off when the dicharge voltage is reached. I have selected a low power comparator (TLV7011) and used a 2.5V LDO to make a stable reference votage. The PCB was just 4.90 USD and components about 2 USD. 

When I received the boards I was happy. Elecrow did a great job.

When I assembled one board I was still happy. I didn't have the battery holder yet, it was still shipping from China.

When I tested the board it didn't work well. BaDiTo V1 FAILED. The discharge current was lower than expected and when connected to a oscilloscope I found oscillation. I realised I made a stupid mistake. The comparator should have been designed with a large hysteresis. When the voltage falls below the threshold, the load is removed, the voltage jumps back up and the load is re-enabled, then the voltage falls again and the load is removed, etc. in an endless loop, which is oscillation. 

Then I remebered this board https://easyeda.com/Iam1CM69/18650-discharge-board designed by https://twitter.com/AstroTogger
First I thought it was a super dumb design, but the longer I tought about it the more brilliant it became. It consist of just power resistors, a $0.10 battery protection module, a LED, and switch. The protection circuit with a DW01 will disconnect the battery when the undervoltage level (2.5V by default) is reached. The DW01 has anti-oscillation built in which is great. The DW01 will only restarted discharging when the load is removed, that is why the switch is there. And the LED is powered from the battery and lit during discharging. When discharging is done the LED turns off and the circuit uses 0.1uA which is super low. I contacted the designer and he said that after the load is disconnected the battery voltage jumps back to almost 2.9V, which is pretty good.