Local Hermes + Qwen3.6:35B agent stalls mid-task announcing actions it never takes
glad-k · reddit · 2026-09-16
A local-AI hobbyist running Hermes with qwen3.6:35b on Ollama (32GB RAM + 8GB VRAM, chosen to fit 128k context) reports that when given a large TODO list, the model works for a while, then announces "And now I'm going to do X:" and simply stops. Telling it to continue produces "Yeah, I'll do it with Y, let's start:" — then it stops again, eventually getting stuck in that loop, with the fake action messages even leaking through to Telegram.
The author posted multiple screenshots and asks whether the model just isn't smart enough for long tasks or whether the setup is misconfigured.
More from coding & agent
- General Reasoning explains how to train agent swarms with RL and latency pressure — rosstaylor90 · 2026-11-02
- "Coding isn't the bottleneck anymore — usage windows are," says founder Kyle Gawley — kylegawley · 2026-11-02
- Demo passed ≠ production-ready: 5 questions to answer before letting agents run your business — _kellhus · 2026-11-02
- An AI agent passed 100 test chats—still not ready to run the business — djgoel123 · 2026-11-02
- Agent given total creative freedom produces a 1:44 code-drawn animated short film — ivegotnoidea1 · 2026-11-02
- Meme: Wake Up, Use Claude Code, Repeat All Day — tekbog · 2026-11-02