[Bug]: GENERATION_FAILED (fp: 69213336)

[mikecheongx]

Description

cant seems to generate a proper downloaded slide file

Steps to reproduce

1. open the app
2. insert prompt: Generate the first 3 slides of a pitch deck for a fintech startup: title slide, problem, solution. 16:9 format, navy palette, one orange accent.
3. error

Expected behavior

a working downloadable ppt should be done

Actual behavior

Error invoking remote method 'codesign:v1:generate': CodesignError: Generation aborted after 1200s (Settings → Advanced → Generation timeout).

open-codesign version

0.1.4

Platform

Windows

OS version

Windows 11 (10.0.26100)

Model provider (if relevant)

None

Error code

GENERATION_FAILED

Diagnostics bundle

Bundle saved locally at ~\Downloads\open-codesign-diagnostics-2026-05-05T20-44-21.zip. Attach it to this issue after submitting.

Logs

toolResultLimit: 8192,
  window: 3
}
[2026-05-06 04:40:45.552] [info]   (chat-messages-ipc) chat.append {
  designId: '88483b5a-9edb-4011-9b16-b08e2061fbb7',
  seq: 113,
  kind: 'assistant_text'
}
[2026-05-06 04:41:20.327] [warn]   (main:ipc) generate.timeout.fired { runId: 'mot2q23t-1ty18g', id: 'mot2q23t-1ty18g', timeoutSec: 1200 }
[2026-05-06 04:41:20.331] [info]   (main:ipc) agent.turn_end {
  runId: 'mot2q23t-1ty18g',
  generationId: 'mot2q23t-1ty18g',
  deltas: 219,
  tools: 0
}
[2026-05-06 04:41:20.332] [info]   (main:ipc) agent.end { runId: 'mot2q23t-1ty18g', generationId: 'mot2q23t-1ty18g' }
[2026-05-06 04:41:20.332] [error]  (main:ipc) [generate] step=send_request.fail {
  runId: 'mot2q23t-1ty18g',
  generationId: 'mot2q23t-1ty18g',
  provider: 'custom-nvidea-i8q4',
  modelId: 'minimaxai/minimax-m2.7',
  ms: 1200014,
  stopReason: 'aborted'
}
[2026-05-06 04:41:20.333] [error]  (main:ipc) generate.fail {
  runId: 'mot2q23t-1ty18g',
  generationId: 'mot2q23t-1ty18g',
  ms: 1200019,
  provider: 'custom-nvidea-i8q4',
  modelId: 'minimaxai/minimax-m2.7',
  baseUrl: 'https://integrate.api.nvidia.com/v1',
  message: 'Generation aborted after 1200s (Settings → Advanced → Generation timeout).',
  code: 'GENERATION_TIMEOUT'
}
[2026-05-06 04:41:20.335] [info]   (chat-messages-ipc) chat.append {
  designId: '88483b5a-9edb-4011-9b16-b08e2061fbb7',
  seq: 114,
  kind: 'assistant_text'
}
[2026-05-06 04:41:20.351] [error]  (renderer:console) [store] applyGenerateError {"generationId":"mot2q23t-1ty18g","designId":"88483b5a-9edb-4011-9b16-b08e2061fbb7","message":"Error invoking remote method 'codesign:v1:generate': CodesignError: Generation aborted after 1200s (Settings → Advanced → Generation timeout)."} {}
[2026-05-06 04:41:20.352] [info]   (chat-messages-ipc) chat.append {
  designId: '88483b5a-9edb-4011-9b16-b08e2061fbb7',
  seq: 115,
  kind: 'error'
}
[2026-05-06 04:41:20.424] [info]   (snapshots-ipc) snapshot.created {
  id: '2672aaac-d98f-4ed4-bfc9-2d0e8a060dcd',
  type: 'edit',
  designId: '88483b5a-9edb-4011-9b16-b08e2061fbb7'
}

[github-actions]

Thanks for the detailed report, @mikecheongx! This is a generation timeout — the model didn't complete its response within the default 1200 s (20 min) limit.

From your logs, the generation ran for exactly 1200 s and was aborted by the timeout timer:

• Provider: custom-nvidea-i8q4
• Model: minimaxai/minimax-m2.7
• Stop reason: aborted

The error message points to Settings → Advanced → Generation timeout. In v0.1.4 (which includes a fix that makes the dropdown work correctly for longer values), you can now set the timeout up to 2 h.

Two things to try:

1. Increase the timeout — Go to Settings → Advanced → Generation timeout and choose a longer value (e.g., 30 min or 1 h). The MiniMax model via NVIDIA may need more time for PPTX generation.
2. Try a different model — If available, test with a faster model (e.g., GPT-4o-mini, Claude Haiku, or a smaller frontier model) to see if the issue is model-specific.

Could you also check whether:

• The model is reachable (Test Connection passes in provider config)?
• The same prompt works with another provider/model (even a small one) within the default timeout?

That would help narrow down whether this is a model latency issue or a bug in the generation pipeline.

---

Open-CoDesign Bot