Training AI to Play Smash Bros with Open Source Models & Custom tmux Tool
yacineMTB · x · 2026-08-07
Developer @yacineMTB shared a hardcore practice of developing agents using open-source models combined with custom tooling.
- Custom Tooling: Leveraging a tmux-auto tool (detailed in a previous blog) to keep complex projects on rails for extended periods.
- Environment Setup: The team extended the Dolphin emulator, pulled public Slippi replays for Super Smash Bros. (Marth vs. Marth), and built a harness to test positions side-by-side.
- Progress: The Marth character logic is nearly complete, with only some edge cases (like simultaneous hit validation) remaining. The rest of the characters are expected to work seamlessly afterward.
More from coding & agent
- DeepMind turns coding agent Antigravity into a science tool, surfacing novel antimicrobial peptide in one session — pushmeet · 2026-10-16
- Cross-review coding workflow: Claude writes the plan, Codex audits it — theteknosaur · 2026-10-16
- Simon Willison: code mode is the future of tool calling, if it can be sandboxed — springrod · 2026-10-16
- Normalizing Claude's schema-free MCP output before it hits downstream n8n nodes — snowingbol · 2026-10-16
- Agent named 'Ilya' reportedly scores 1.1pt higher on benchmarks — daniel_mac8 · 2026-10-16
- A plain notes.md fixed my agent's cross-session amnesia better than chat history — rahul_watertech · 2026-10-16