opencode stucks while trying to write a file

[vldF]

When I try to generate a code, the opencode can read files and even run bash commands. However, when it tries to write a file with tool (I guess), it hangs for some time and then TimeoutError: The operation timed out. appears:

I have tried multiple models, now, I try deepseek r1. Also, I use custom proxy with openai/openrouter interface, but it works fine with continue. How can I troubleshoot it? Logs contains nothing

2025-08-21T131406.log

[vldF]

Hmm, it can create AGENT.md file when I do /init. Probably it's something with LSP? I have no one locally

[vldF]

I disabled golsp (is isn't on my computer, whatever)

"lsp": {
    "gopls": {
        "disabled": true
    }

Nothing changed: it can't write .go files and gets stuck on preparing write... step

[rekram1-node]

very strange:

ERROR 2025-08-21T13:19:06 +84ms service=session session=ses_733e88c4effePUe2uI4BP5zG9I error={}
INFO  2025-08-21T13:19:06 +0ms service=bus type=session.error publishing
DEBUG 2025-08-21T13:19:06 +2ms service=tui timeTakenMs=2 messages.renderView
ERROR 2025-08-21T13:19:06 +1ms service=tui message=TimeoutError: The operation timed out. name=UnknownError Server error
INFO  2025-08-21T13:19:06 +1ms service=bus type=storage.write publishing
INFO  2025-08-21T13:19:06 +0ms service=bus type=message.part.updated publishing
INFO  2025-08-21T13:19:06 +0ms service=bus type=storage.write publishing
INFO  2025-08-21T13:19:06 +0ms service=bus type=message.updated publishing
INFO  2025-08-21T13:19:06 +1ms service=session session=ses_733e88c4effePUe2uI4BP5zG9I sessionID=ses_733e88c4effePUe2uI4BP5zG9I unlocking
INFO  2025-08-21T13:19:06 +0ms service=bus type=session.idle publishing
INFO  2025-08-21T13:19:06 +1ms service=server duration=293672 response
DEBUG 2025-08-21T13:19:06 +2ms service=tui message=msg_8ccc3feb8001glAwR7dkfMZk88 part=prt_8ccc424c6001hEtSrSv3r1VUjw message part updated
DEBUG 2025-08-21T13:19:06 +3ms service=tui pending render, skipping

disabled_providers=["openai","gemini","openrouter"] small_model=ollama/gemma3:1b provider={"ollama":{"name":"Ollama (local)","npm":"@ai-sdk/openai-compatible","models":{"gemma3:1b":{"name":"gemma3:1b"},"gpt-oss:20b":{"name":"gpt-oss:20b"}},"options":{"baseURL":"http://localhost:11434/v1"}},"llm-proxy":{"name":"<my-provider-name>","npm":"@ai-sdk/openai-compatible","models":{"z-ai/glm-4.5":{"name":"z-ai/glm-4.5"},"qwen3-coder-plus":{"name":"qwen3-coder-plus"},"moonshotai/kimi-k2":{"name":"kimi-k2"},"deepseek-reasoner":{"name":"deepseek-reasoner r1"},"deepseek-chat":{"name":"deepseek-chat v3"}},"options":{"baseURL":"<my-base-url>/"}}} permission={"edit":"allow","bash":"allow","webfetch":"allow"} agent={} mode={} plugin=[] username=v.feofilaktov loaded

what provider are you using for deepseek-reasoner?

[vldF]

@rekram1-node, I use my custom company's provider with "npm": "@ai-sdk/openai-compatible" set

[rekram1-node]

[automated] Closing due to 90+ days of inactivity. Feel free to reopen if you still need this!