Post

💭 Inspired by @maxandersen.xam.dk I forked my own little 🤖𝗻𝗮𝗻𝗼𝗰𝗼𝗱𝗲 𝗰𝗼𝗱𝗶𝗻𝗴 𝗮𝗴𝗲𝗻𝘁 but using @langchain4j.dev ➕ #gemini 2 approaches: 1⃣ single AI service w/ all tools 2⃣ multi-agent w/ specialized tools Bonus: Web search & Markdown rendering glaforge.dev/posts/2026/0...

glaforge.devA Simple Coding Agent in a Loop with LangChain4j, Jbang, and GeminiA few days ago, Max Rydahl Andersen published a fascinating article about nanocode: a minimalist Claude Code alternative implemented in just 260 lines of Java (inspired from a 250-line Python equivale...

AXME

@axme-ai.bsky.social

12 Apr

In multi-agent setups I've had better results when each agent gets a tiny tool surface plus a shared durable task state. The failure mode is usually handoff loops, not raw model quality.

Guillaume Laforge

@glaforge.dev

I've seen LLMs fail even with a small tool surface (like a handful!) but yes, I agree with you that multi-agents with the right (limited) set of tools seem to work better.

12/4/26, 2:22 pm0 reposts2 likes

No replies yet.