Seth
827850b8d7
- IDEA.md: project scope (Minecraft ops AI assistant via qwen3-coder LoRA/SFT) - PLAN.md: complete roadmap with prior art analysis, architecture, phased plan, dev server docs - data/schema.json: training example JSON Schema with negative_output support - data/processed/seed_dataset.jsonl: 31 validated examples from repair code, prayer logs, session history - data/validate_dataset.py: schema validator with summary statistics - ingame/: Mineflayer bot framework (test_connect, spawn_bots, aware_bots with full event logging) - Directory structure for knowledge/, eval/, training/, agent/ (Phase 1.3+ work)
16 lines
1.1 KiB
Markdown
16 lines
1.1 KiB
Markdown
# IDEA.md -- Project Idea
|
|
|
|
## What is this?
|
|
A lightweight, Minecraft-specific AI assistant built by adapting `qwen3-coder` (not training from scratch). It should act as an operations copilot for Sethpc Minecraft servers: generating correct commands, troubleshooting logs, and safely automating repetitive admin tasks.
|
|
|
|
## Problem it solves
|
|
General-purpose models carry broad knowledge and can feel heavy, expensive, and inconsistent for Minecraft server operations. This project narrows focus to Minecraft workflows so responses are faster, cheaper, and more accurate for tasks like RCON command generation, event parsing, world border automation, and 1.21 syntax correctness.
|
|
|
|
## Constraints / preferences
|
|
- Base model: `qwen3-coder` (adapt/fine-tune with LoRA/SFT)
|
|
- Scope: Minecraft ops assistant (not a full new foundation model)
|
|
- Prioritize tool use + guardrails over memorization
|
|
- Must understand local homelab server context and documented workflows
|
|
- Prefer iterative approach: baseline -> dataset -> fine-tune -> eval -> deploy
|
|
- Safety: command allowlists, destructive-action confirmations, auditable logs
|