本地模型完全没法儿用了. 能让我回到以前的版本吗?

[linzhanyu]

Description

{"message":"Failed to parse input at pos 388: <tool_call>\n<function=write>\n\n</tool_call>","type":"server_error","code":500}

第一次几乎都在报这个错.
omo 的 /start-work 由于这个错误基本就循环卡死了.

我不想自动更新!
这个问题有几天了, 大概是 1.18.10 开始? 还是1.10.18开始的.

Plugins

oh-my-openagent

OpenCode version

1.18.18

Steps to reproduce

No response

Screenshot and/or share link

No response

Operating System

Windows 10 & Ubuntu 22

Terminal

No response

[github-actions]

This issue might be a duplicate of existing issues. Please check:

• Progress halts with qwen 3.6 35b-a3b with naked tool call in the console #24316: Local model (Qwen/llama.cpp) outputs raw XML <tool_call> tags instead of JSON, causing parse failures and the session to halt — the error in your report (Failed to parse input at pos 388: <tool_call>\n<function=write>) matches this pattern exactly.
• Invalid-tool-call error text echoes provider special tokens back into history, causing cascading tool-call failures (Kimi K2 on vLLM providers) #37297: Invalid tool-call parse errors with local/vLLM providers cause cascading failures and loops — similar symptom of a session getting stuck after a parse error.

If your local model is a Qwen or similar model running via llama.cpp/Ollama, #24316 is likely the same issue.

[linzhanyu]

it's duplicated.