[Bug]: GENERATION_FAILED (fp: c85d010d)

[Maxitendo1]

Description

Error invoking remote method 'codesign:v1:generate': CodesignError: terminated

OS: Kubuntu 25.10
AI Provider: Lemonade with OpenAI Chat
Model: Qwen3.5-9B-Uncensored-HauhauCS-Aggressive
Connection: OK

Steps to reproduce

1. Open the app
2. Start a new design
3. Type something

Expected behavior

It should work.

Actual behavior

Error invoking remote method 'codesign:v1:generate': CodesignError: terminated

open-codesign version

0.1.4

Platform

Linux

OS version

6.17.0-22-generic

Model provider (if relevant)

Other

Error code

GENERATION_FAILED

Diagnostics bundle

Bundle saved locally at ~

open-codesign-diagnostics-2026-04-26T09-46-02.zip

/Downloads/open-codesign-diagnostics-2026-04-26T09-46-02.zip. Attach it to this issue after submitting.

Logs

at writeOrBuffer (node:internal/streams/writable:572:12)
    at _write (node:internal/streams/writable:501:10)
    at Writable.write (node:internal/streams/writable:510:10)
    at console.value (node:internal/console/constructor:303:16)
    at console.error (node:internal/console/constructor:433:26)
    at transport.writeFn ([url omitted] omitted]
[2026-04-26 11:46:02.845] [error]  [crash:browser] Error: write EPIPE
    at afterWriteDispatched (node:internal/stream_base_commons:159:15)
    at writeGeneric (node:internal/stream_base_commons:150:3)
    at Socket._writeGeneric (node:net:966:11)
    at Socket._write (node:net:978:8)
    at writeOrBuffer (node:internal/streams/writable:572:12)
    at _write (node:internal/streams/writable:501:10)
    at Writable.write (node:internal/streams/writable:510:10)
    at console.value (node:internal/console/constructor:303:16)
    at console.error (node:internal/console/constructor:433:26)
    at transport.writeFn ([url omitted] omitted]
[2026-04-26 11:46:02.845] [error]  Unhandled Error: write EPIPE
    at afterWriteDispatched (node:internal/stream_base_commons:159:15)
    at writeGeneric (node:internal/stream_base_commons:150:3)
    at Socket._writeGeneric (node:net:966:11)
    at Socket._write (node:net:978:8)
    at writeOrBuffer (node:internal/streams/writable:572:12)
    at _write (node:internal/streams/writable:501:10)
    at Writable.write (node:internal/streams/writable:510:10)
    at console.value (node:internal/console/constructor:303:16)
    at console.error (node:internal/console/constructor:433:26)
    at transport.writeFn ([url omitted] omitted]
[2026-04-26 11:46:02.845] [error]  [crash:browser] Error: write EPIPE
    at afterWriteDispatched (node:internal/stream_base_commons:159:15)
    at writeGeneric (node:internal/stream_base_commons:150:3)
    at Socket._writeGeneric (node:net:966:11)
    at Socket._write (node:net:978:8)
    at writeOrBuffer (node:internal/streams/writable:572:12)
    at _write (node:internal/streams/writable:501:10)
    at Writable.write (node:internal/streams/writable:510:10)
    at console.value (node:internal/console/constructor:303:16)
    at console.error (node:internal/console/constructor:433:26)
    at transport.writeFn ([url omitted] omitted]
[2026-04-26 11:46:02.845] [error]  Unhandled Error: write EPIPE
    at afterWriteDispatched (node:internal/stream_base_commons:159:15)
    at writeGeneric (node:internal/stream_base_commons:150:3)
    at Socket._writeGeneric (node:net:966:11)
    at Socket._write (node:net:978:8)
    at writeOrBuffer (node:internal/streams/writable:572:12)
    at _write (node:internal/streams/writable:501:10)
    at Writable.write (node:internal/streams/writable:510:10)
    at console.value (node:internal/console/constructor:303:16)
    at console.error (node:internal/console/constructor:433:26)
    at transport.writeFn ([url omitted] omitted]

[github-actions]

Thanks for reporting this. I checked the current 0.1.4 code path and what I can verify is narrower than the symptom suggests.

Right now, Open CoDesign only turns a raw terminated transport error into a specific relay hint when the request used openai-responses against a custom base URL. If the request was sent over openai-chat, the same terminated shape falls through to unknown, and the renderer suppresses that generic hypothesis, so the UI ends up showing only CodesignError: terminated. That matches existing issue #189, so this looks related rather than a new classified error path. I also found #216 about test-connection vs runtime-call parity, but I did not find a matching PR for this exact terminated path.

Based on the model ID you shared, this does not look like the developer-role regression from #183 / #234: third-party openai-chat reasoning is only enabled for a small allowlist, and Qwen3.5-9B-Uncensored-HauhauCS-Aggressive would not match it.

Relevant code:

• packages/shared/src/diagnostics.ts:219 — only openai-responses + custom baseUrl + terminated/truncated-stream patterns get the relayStreamingBug hint.
• packages/shared/src/diagnostics.ts:270 — without HTTP status, only a limited set of message patterns are upgraded; otherwise it falls back to unknown.
• apps/desktop/src/renderer/src/store.ts:1100 — the renderer drops the bare unknown hypothesis, so the user just sees the upstream error text.
• packages/providers/src/wire-policy.ts:50 — third-party openai-chat reasoning is gated by a narrow model-ID allowlist.

Need more info:

• Please attach the diagnostics bundle to the issue, or paste the redacted generate.fail block from the main log (provider, modelId, baseUrl, status, message). The snippet here only shows write EPIPE crash/browser output.
• Please confirm the exact custom provider settings: selected wire (OpenAI Chat vs OpenAI Responses) and the redacted base URL shape.

---

open-codesign Bot

[github-actions]

Thanks for the detailed report and for attaching the diagnostics bundle, Maxi. The previous bot analysis already covered the code path for CodesignError: terminated, but I've reviewed the attached bundle and your logs more closely.

Your log shows repeated write EPIPE errors, which indicate the provider connection was broken before the generation could complete. This is consistent with the terminated error from the OpenAI-compatible relay. Since you're using OpenAI Chat wire (not OpenAI Responses), the current diagnostics code at packages/shared/src/diagnostics.ts:219 does not upgrade this to a specific hint — it falls through to the generic unknown bucket, which the renderer suppresses.

Key observations:

• You're running Qwen3.5-9B via Lemonade (a local provider). The write EPIPE suggests the local model process (or Lemonade's bridge) crashed or timed out during streaming.
• packages/providers/src/wire-policy.ts:50 shows that third-party openai-chat reasoning is only enabled for a narrow allowlist; your model likely does not match, but that would cause a different error, not terminated.
• The terminated error in the OpenAI-compatible relay (packages/providers/src/relay.ts around line 180) is raised when the fetch stream ends unexpectedly (e.g., connection reset by the provider).

What would help narrow this down:

1. Which Lemonade backend are you using (Ollama, llama.cpp, vLLM)? Does the Qwen3.5-9B model start successfully when tested via curl or the Lemonade UI?
2. Does the model support streaming via OpenAI-compatible chat completions? Some local backends require explicit "stream": true or have different endpoint paths.
3. Please share the redacted generate.fail block from the main log (around the time of failure, the logger writes provider, modelId, baseUrl, status, and message). The diagnostics bundle may contain this; you can paste it here if it's not sensitive.
4. What timeout / max tokens settings do you have configured? A 9B model may take >60 seconds for the first token, and the default timeout might be too short.

Let's get this sorted!

Open-CoDesign Bot

[Maxitendo1]

1. Lemonade backend: llama.cpp, qwen3.5 model is starting succesfully in the lemonade ui. i downloaded the model from huggingface:

https://huggingface.co/HauhauCS/Qwen3.5-9B-Uncensored-HauhauCS-Aggressive

2. I don't know.
3. [2026-05-01 22:47:15.693] [error] (main:ipc) generate.fail {
   runId: 'mondvtyx-1cmhna',
   generationId: 'mondvtyx-1cmhna',
   ms: 7242,
   provider: 'custom-lemonade-v70l',
   modelId: 'user.Qwen3.5 Unzensiert',
   baseUrl: '[url omitted]',
   message: 'terminated',
   code: 'PROVIDER_ERROR'
   }
   [2026-05-01 22:47:15.700] [error] (renderer:console) [store] applyGenerateError {"generationId":"mondvtyx-1cmhna","designId":"a9711cd4-a433-45c6-8168-15b787075699","message":"Error invoking remote method 'codesign:v1:generate': CodesignError: terminated"} {}
   [2026-05-01 22:47:15.701] [info] (chat-messages-ipc) chat.append {
   designId: 'a9711cd4-a433-45c6-8168-15b787075699',
   seq: 1,
   kind: 'error'
   }
   [2026-05-01 22:47:20.685] [info] (main:ipc) [title] step=send_request.ok { runId: 'dc9d9d22-5673-43f2-88a3-550cd7752d43', ms: 12266 }
   [2026-05-01 22:47:20.685] [error] (main:ipc) [title] step=send_request.fail {
   runId: 'dc9d9d22-5673-43f2-88a3-550cd7752d43',
   ms: 12266,
   errorClass: 'CodesignError'
   }
   [2026-05-01 22:47:20.685] [error] (main:ipc) [title] generate-title.fail {
   runId: 'dc9d9d22-5673-43f2-88a3-550cd7752d43',
   provider: 'custom-lemonade-v70l',
   modelId: 'user.Qwen3.5 Unzensiert',
   baseUrl: '[url omitted]',
   message: 'Model returned empty title',
   code: 'PROVIDER_ERROR'
   }
   [2026-05-01 22:47:20.686] [warn] (renderer:store) [title] generateTitle failed, using prompt fallback {
   designId: 'a9711cd4-a433-45c6-8168-15b787075699',
   message: "Error invoking remote method 'codesign:v1:generate-title': CodesignError: Model returned empty title"
   }
   [2026-05-01 22:47:20.687] [info] (snapshots-ipc) design.renamed {
   id: 'a9711cd4-a433-45c6-8168-15b787075699',
   name: 'Build a landing page for an AI startup.…'
   }
   [2026-05-01 22:47:28.842] [error] (main:updates) autoUpdater.error {
   message: 'Cannot find latest-linux.yml in the latest release artifacts ([url omitted] HttpError: 404 \n' +
   '"method: GET url: [url omitted] omitted] double check that your authentication token is correct. Due to security reasons, actual status maybe not reported, but 404.[path omitted]"\n' +
   'Headers: {\n' +
   ' "cache-control": "no-cache",\n' +
   ' "content-length": "9",\n' +
   "content-security-policy": "default-src 'none'; base-uri 'self'; connect-src 'self'; form-action 'self'; img-src 'self' data:; script-src 'self'; style-src 'unsafe-inline'",\n +
   ' "content-type": "text/plain; charset=utf-8",\n' +
   ' "date": "Fri, 01 May 2026 20:47:28 GMT",\n' +
   ' "referrer-policy": "no-referrer-when-downgrade",\n' +
   ' "server": "github.com",\n' +
   ' "strict-transport-security": "max-age=31536000; includeSubdomains; preload",\n' +
   ' "vary": "X-PJAX, X-PJAX-Container, Turbo-Visit, Turbo-Frame, X-Requested-With, Sec-Fetch-Site,Accept-Encoding, Accept, X-Requested-With",\n' +
   ' "x-content-type-options": "nosniff",\n' +
   ' "x-frame-options": "deny",\n' +
   ' "x-github-request-id": "A49A:1702:27644D:1D11CB:69F51160",\n' +
   ' "x-xss-protection": "0"\n' +
   '}\n' +
   ' at createHttpError ([url omitted] omitted]' +
   ' at ElectronHttpExecutor.handleResponse ([url omitted] omitted]' +
   ' at ClientRequest. ([url omitted] omitted]' +
   ' at ClientRequest.emit (node:events:519:28)\n' +
   ' at SimpleURLLoaderWrapper. (node:electron/js2c/browser_init:2:130549)\n' +
   ' at SimpleURLLoaderWrapper.emit (node:events:519:28)\n' +
   ' at SimpleURLLoaderWrapper.callbackTrampoline (node:internal/async_hooks:130:17)',
   stack: 'Error: Cannot find latest-linux.yml in the latest release artifacts ([url omitted] HttpError: 404 \n' +
   '"method: GET url: [url omitted] omitted] double check that your authentication token is correct. Due to security reasons, actual status maybe not reported, but 404.[path omitted]"\n' +
   'Headers: {\n' +
   ' "cache-control": "no-cache",\n' +
   ' "content-length": "9",\n' +
   "content-security-policy": "default-src 'none'; base-uri 'self'; connect-src 'self'; form-action 'self'; img-src 'self' data:; script-src 'self'; style-src 'unsafe-inline'",\n +
   ' "content-type": "text/plain; charset=utf-8",\n' +
   ' "date": "Fri, 01 May 2026 20:47:28 GMT",\n' +
   ' "referrer-policy": "no-referrer-when-downgrade",\n' +
   ' "server": "github.com",\n' +
   ' "strict-transport-security": "max-age=31536000; includeSubdomains; preload",\n' +
   ' "vary": "X-PJAX, X-PJAX-Container, Turbo-Visit, Turbo-Frame, X-Requested-With, Sec-Fetch-Site,Accept-Encoding, Accept, X-Requested-With",\n' +
   ' "x-content-type-options": "nosniff",\n' +
   ' "x-frame-options": "deny",\n' +
   ' "x-github-request-id": "A49A:1702:27644D:1D11CB:69F51160",\n' +
   ' "x-xss-protection": "0"\n' +
   '}\n' +
   ' at createHttpError ([url omitted] omitted]' +
   ' at ElectronHttpExecutor.handleResponse ([url omitted] omitted]' +
   ' at ClientRequest. ([url omitted] omitted]' +
   ' at ClientRequest.emit (node:events:519:28)\n' +
   ' at SimpleURLLoaderWrapper. (node:electron/js2c/browser_init:2:130549)\n' +
   ' at SimpleURLLoaderWrapper.emit (node:events:519:28)\n' +
   ' at SimpleURLLoaderWrapper.callbackTrampoline (node:internal/async_hooks:130:17)\n' +
   ' at newError ([url omitted] omitted]' +
   ' at fetchData ([url omitted] omitted]' +
   ' at async GitHubProvider.getLatestVersion ([url omitted] omitted]' +
   ' at async AppImageUpdater.getUpdateInfoAndProvider ([url omitted] omitted]' +
   ' at async AppImageUpdater.doCheckForUpdates ([url omitted] omitted]'
   }
   [2026-05-01 22:47:28.843] [error] (main:updates) startup.checkForUpdates.fail {
   message: 'Cannot find latest-linux.yml in the latest release artifacts ([url omitted] HttpError: 404 \n' +
   '"method: GET url: [url omitted] omitted] double check that your authentication token is correct. Due to security reasons, actual status maybe not reported, but 404.[path omitted]"\n' +
   'Headers: {\n' +
   ' "cache-control": "no-cache",\n' +
   ' "content-length": "9",\n' +
   "content-security-policy": "default-src 'none'; base-uri 'self'; connect-src 'self'; form-action 'self'; img-src 'self' data:; script-src 'self'; style-src 'unsafe-inline'",\n +
   ' "content-type": "text/plain; charset=utf-8",\n' +
   ' "date": "Fri, 01 May 2026 20:47:28 GMT",\n' +
   ' "referrer-policy": "no-referrer-when-downgrade",\n' +
   ' "server": "github.com",\n' +
   ' "strict-transport-security": "max-age=31536000; includeSubdomains; preload",\n' +
   ' "vary": "X-PJAX, X-PJAX-Container, Turbo-Visit, Turbo-Frame, X-Requested-With, Sec-Fetch-Site,Accept-Encoding, Accept, X-Requested-With",\n' +
   ' "x-content-type-options": "nosniff",\n' +
   ' "x-frame-options": "deny",\n' +
   ' "x-github-request-id": "A49A:1702:27644D:1D11CB:69F51160",\n' +
   ' "x-xss-protection": "0"\n' +
   '}\n' +
   ' at createHttpError ([url omitted] omitted]' +
   ' at ElectronHttpExecutor.handleResponse ([url omitted] omitted]' +
   ' at ClientRequest. ([url omitted] omitted]' +
   ' at ClientRequest.emit (node:events:519:28)\n' +
   ' at SimpleURLLoaderWrapper. (node:electron/js2c/browser_init:2:130549)\n' +
   ' at SimpleURLLoaderWrapper.emit (node:events:519:28)\n' +
   ' at SimpleURLLoaderWrapper.callbackTrampoline (node:internal/async_hooks:130:17)'
   }
4. I haven't configured anything, I'm just using it as is.

[skarone]

I think is llama.cpp, I'm using qwen3.5-q8 9b and is also failing with the same error.
I tested the same model using ollama, and I don't get the error but the it thinks the first time then stops responding

[Sun-sunshine06]

Closing this repeated terminated / transport-interrupted report in favor of #314 after the v0.2 provider diagnostics sweep. Current main includes the provider failure diagnostics bridge and the #315 diagnostics/recovery cleanup; the upcoming v0.2 package will include them.

If this still reproduces on the v0.2 package, please open or reopen with a fresh diagnostics bundle and link #314 so we can keep the remaining provider/gateway follow-up in one place.