What causes "${message.errorMessage ?? "Provider error"}" in can1357/oh-my-pi?

Inspect the thrown errorMessage for the specific provider cause

can1357/oh-my-pi · error · Error

${message.errorMessage ?? "Provider error"}

Error message

${message.errorMessage ?? "Provider error"}

What it means

The inference complete() method streams a model message and, if message.stopReason === "error", throws with the provider's errorMessage (defaulting to "Provider error"). This surfaces upstream LLM failures (auth, network, quota, malformed request) from the conventional-commit inference layer, where results are also cached only on success.

Source

Thrown at packages/coding-agent/src/commit/conventional/inference.ts:132

			try {
				this.#onProgress?.(request.progressLabel);
				const timeout = AbortSignal.timeout(120_000);
				const signal = this.#signal ? AbortSignal.any([this.#signal, timeout]) : timeout;
				const message = await completeSimple(
					target.model,
					{
						systemPrompt: request.systemPrompt.trim() ? [request.systemPrompt] : undefined,
						messages: [{ role: "user", content: request.userPrompt, timestamp: Date.now() }],
					},
					{
						apiKey: target.apiKey,
						maxTokens: 16_384,
						reasoning,
						signal,
					},
				);
				responseText = extractAssistantText(message);
				if (message.stopReason === "error") throw new Error(message.errorMessage ?? "Provider error");
				if (!responseText.trim()) throw new Error("Empty model response");
				const raw = { text: responseText, stopReason: message.stopReason };
				const parsed = parse(raw);
				if (request.cacheable !== false) {
					this.#cache?.put({
						key,
						model: modelKey,
						operation: request.operation,
						request: requestJson,
						response: {
							text: responseText,
							stopReason: message.stopReason,
							costUsd: message.usage.cost.total,
						},
					});
				}
				this.#cache?.recordUsage(modelKey, request.operation, message.usage);
				return parsed;

View on GitHub (pinned to 9690622007)

Solutions

Inspect the thrown errorMessage for the specific provider cause
Verify provider credentials and the configured model id
Retry on transient network/rate-limit errors; consider the built-in cache which avoids repeat calls
Reduce input size (smaller diff) if the error indicates context/token limits

Example fix

// before (no key in env)
omp commit  # -> Provider error: 401 unauthorized
// after
export ANTHROPIC_API_KEY=sk-... && omp commit

Defensive patterns

Strategy: retry

Validate before calling

if (!process.env.ANTHROPIC_API_KEY && !process.env.OPENAI_API_KEY) {
  throw new Error("No provider credentials configured for inference");
}

Try / catch

try {
  const result = await inference.complete(request);
} catch (err) {
  if (err.message === "Provider error" || /rate limit|timeout|5\d\d/i.test(err.message)) {
    await Bun.sleep(backoffMs);
    return inference.complete(request); // bounded retries
  }
  throw err;
}

Prevention

Configure and rotate provider API keys before automation runs
Use bounded exponential backoff for transient provider failures
Keep prompts (diffs) within model context limits
Rely on the inference cache to avoid redundant paid calls

When it happens

Trigger: Any complete() call whose underlying provider request ends in an error stop reason: invalid API key, unreachable endpoint, model id not found, context window exceeded (maxTokens: 16384 plus prompt), or provider-side 5xx.

Common situations: Missing/expired credentials in the environment; offline or proxied networks blocking the provider host; requesting a model the account cannot access; oversized diffs pushing the prompt past limits.

Related errors

AI-assisted analysis of can1357/oh-my-pi@9690622007 (2026-08-31). Data as JSON: /api/errors/56def090fefc6aea. Report an issue: GitHub.