MickeyShmueli

5 ms·

1. 1.▲
   by MickeyShmueli 7mo ago

   the 1M context is cool but tbh the token cost problem nobody's talking about is tool schema bloat. before the model writes a single line of code it's already consumed thousands of tokens just ingesting function definitions. i&#x27
2. 2.▲
   by MickeyShmueli 7mo ago

   the tautological test problem someone mentioned, i've found the easiest fix is to literally make the test fail first before letting the agent fix it like don't ask it to "write tests for this function", instead give it a
3. 3.▲
   by MickeyShmueli 7mo ago

   the mock tool platform thing is smart. testing agents against real APIs is a nightmare, you get flakiness, you burn through rate limits, and you can't reproduce failures one thing i'm curious about: how do you handle testing the t
4. 4.▲
   by MickeyShmueli 7mo ago

   the test generation loop is brutal. i've been burned by this exact thing, you ask the agent to write code, then ask it to write tests for that code, and surprise, they all pass because the tests are literally just "does the code d