Kiwi-chan's Log: Birch, Boredom, and Bug Fixes! 🥝

Dev.to / 4/30/2026

💬 OpinionDeveloper Stack & InfrastructureIdeas & Deep AnalysisTools & Practical Usage

Key Points

  • Kiwi-chan is making incremental progress in a Minecraft survival loop (get wood, build a base), but repeatedly fails to gather oak logs and then switches to birch logs due to the “OAK OBSESSION BAN” rule.
  • The system shows signs of environmental adaptation by responding to repeated oak-gathering failures with a fallback strategy, though it also triggers a “BOREDOM TRIGGERED” alert and needs a rule-based Coach to diversify tasks.
  • Debugging remains a major bottleneck: frequent “Code extraction failed” errors and the model’s tendency to generate syntactically incorrect JavaScript are causing ongoing manual fixes.
  • Audit and safety rules are helping reduce failures, with before/after item count checks catching pickup issues and pathfinding rules preventing the agent from getting stuck.
  • The author frames the work as building toward a more autonomous Minecraft AI and asks for community support to upgrade GPU hardware for better performance and more complex behaviors.

Another four hours down in the Minecraft world with Kiwi-chan, and it's been a steady, if slightly repetitive, climb! We're still firmly focused on the core survival loop: get wood, build a base. However, Kiwi-chan is very insistent on finding oak logs, even when they aren't around. The "OAK OBSESSION BAN" rule is getting a workout, forcing the AI to switch to birch logs when oak proves elusive.

The logs show a fascinating pattern of repeated failures with gather_oak_log, followed by a switch to gather_birch_log. It's a good sign that the system is responding to the environment and attempting to adapt. We've also hit a "BOREDOM TRIGGERED" alert – apparently, repeatedly digging dirt gets old, even for an AI! The Coach (that's our rule-based decision-making system) is stepping in to diversify tasks.

Debugging continues to be a major focus. We're seeing a lot of "Code extraction failed" errors, and the AI is struggling with basic tasks like digging dirt. I've been manually fixing the generated code, but it's a constant battle. It's clear the LLM is still prone to generating syntactically incorrect JavaScript.

The audit rules are proving invaluable. The beforeCount and afterCount checks are catching failed item pickups, and the pathfinding rules are helping to avoid getting stuck. I've also reinforced the importance of finding open ground for block placement and avoiding hardcoded coordinates.

Overall, progress is incremental, but the system is learning. The constant cycle of failure, debugging, and adaptation is exactly what we need to build a truly autonomous Minecraft AI. It's a bit like building Frankenstein's monster, honestly – lots of patching and hoping it doesn't turn on me! 😅

Help Kiwi-chan Level Up! 🥝✨

This project is fueled by passion (and a rapidly overheating GPU!). Any contribution, big or small, directly helps me upgrade my melting "Frankenstein" rig to an RTX 3060, which will dramatically improve Kiwi-chan's processing power and allow for more complex behaviors.