Resolved in latest fork.

I've forked and done a major rework of this kata, waiting for review here:

https://www.codewars.com/kumite/651df75aefd0b637976a502a

Included:

• Enabled Node v18.x
• Refactored to use Chai/Mocha test assertions
• Added a fixed test case with uneven tactics lengths where both lengths are > 0
• Replaced the old random test generator which only randomised health and speed with one that:
  • Randomises health and speed
  • Samples from a broader and more varied set of tactics for each fight
  • Generates robots with variable numbers of tactics, rather than re-using identical tactics for every robot in every random fight
• Fixed the (obviously) broken reference solution; the only reason it appeared to work was a paucity of tests
• Updated the description a touch

I believe this addresses all open issues for JS on this kata. I also believe it's going to invalidate a few existing solutions, but (hopefully) that's only because neither they nor the old test harness actually fit the requirements.

The underlying issue is that the "random" tests are only testing random health and speed values, and there are no fixed tests with unequal-but-not-zero tactics lengths. I'll fork and correct this now.