vxcontrol/pentagi · error

failed to refine subtasks for task %d: %w

Error message

failed to refine subtasks for task %d: %w

What it means

After loading existing subtasks, RefineSubtasks calls the LLM provider's RefineSubtasks to generate a new plan. This error wraps a failure of that provider call — network errors, provider API errors, malformed LLM output, or missing provider credentials. Refinement aborts before any DB mutation.

Source

Thrown at backend/pkg/controller/subtasks.go:105

			Description: info.Description,
		})
		if err != nil {
			return fmt.Errorf("failed to create subtask for task %d: %w", stc.taskCtx.TaskID, err)
		}
	}

	return nil
}

func (stc *subtaskController) RefineSubtasks(ctx context.Context) error {
	subtasks, err := stc.taskCtx.DB.GetTaskSubtasks(ctx, stc.taskCtx.TaskID)
	if err != nil {
		return fmt.Errorf("failed to get task %d subtasks: %w", stc.taskCtx.TaskID, err)
	}

	plan, err := stc.taskCtx.Provider.RefineSubtasks(ctx, stc.taskCtx.TaskID)
	if err != nil {
		return fmt.Errorf("failed to refine subtasks for task %d: %w", stc.taskCtx.TaskID, err)
	}

	if len(plan) == 0 {
		return nil // no subtasks refined
	}

	subtaskIDs := make([]int64, 0, len(subtasks))
	for _, subtask := range subtasks {
		if subtask.Status == database.SubtaskStatusCreated {
			subtaskIDs = append(subtaskIDs, subtask.ID)
		}
	}

	err = stc.taskCtx.DB.DeleteSubtasks(ctx, subtaskIDs)
	if err != nil {
		return fmt.Errorf("failed to delete subtasks for task %d: %w", stc.taskCtx.TaskID, err)
	}
	stc.InvalidateSubtasks(subtaskIDs)

View on GitHub (pinned to ea665308ba)

Solutions

1. Inspect the wrapped cause in logs to identify provider vs network failure
2. Verify provider API key and base URL in settings/.env
3. Retry the refine operation; LLM calls are idempotent up to this point
4. Configure a fallback provider in provider config

if provider == nil { return fmt.Errorf("no provider configured") }
// pre-check credentials with a lightweight provider health/list-models call

plan, err := stc.taskCtx.Provider.RefineSubtasks(ctx, taskID)
if err != nil {
    if errors.Is(err, context.DeadlineExceeded) { return retryWithBackoff(err) }
    return fmt.Errorf("failed to refine subtasks for task %d: %w", taskID, err)
}

When it happens

Trigger: The configured LLM provider is unreachable, the API key is invalid/quota-exhausted, the model returns unparseable JSON, or ctx is cancelled during the (long) LLM call.

Common situations: Expired or missing provider API key in .env; provider rate limit hit during heavy usage; model name changed after provider upgrade; request timeout on slow models.

Related errors

• failed to switch provider: %w
• failed to perform agent chain for subtask %d: %w
• failed to generate subtasks for task %d: %w
• failed to get task title: %w
• failed to test provider: %w

AI-assisted analysis of vxcontrol/pentagi@ea665308ba (2026-09-01). Data as JSON: /api/errors/e35e1fb09d1f0076. Report an issue: GitHub.