n8n-io/n8n · error · Error

LangSmith client not initialized - check LANGSMITH_API_KEY

Error message

LangSmith client not initialized - check LANGSMITH_API_KEY

What it means

Thrown when `--backend langsmith` is requested but `setupTestEnvironment(...)` did not produce an `lsClient`. The client is built from the `LANGSMITH_API_KEY` environment variable; if it is absent or empty the client is `undefined` and the run aborts before touching the LangSmith API. This is a deliberate early-validation so the failure points at credentials rather than at a downstream 'unauthorized' response.

Source

Thrown at packages/@n8n/ai-workflow-builder.ee/evaluations/cli/index.ts:485

 */
export async function runV2Evaluation(): Promise<void> {
	const args = parseEvaluationArgs();

	if (args.backend === 'langsmith' && (args.prompt || args.promptsCsv || args.testCase)) {
		throw new Error(
			'LangSmith mode requires `--dataset` and does not support `--prompt`, `--prompts-csv`, or `--test-case`',
		);
	}

	// Setup environment with per-stage model configuration
	const logger = createLogger(args.verbose);
	const stageModels = argsToStageModels(args);

	const env = await setupTestEnvironment(stageModels, logger);

	// Validate LangSmith client early if langsmith backend is requested
	if (args.backend === 'langsmith' && !env.lsClient) {
		throw new Error('LangSmith client not initialized - check LANGSMITH_API_KEY');
	}

	// Create workflow generator based on agent type
	const generateWorkflow =
		args.agent === AGENT_TYPES.CODE_BUILDER
			? createCodeWorkflowBuilderGenerator(
					env.parsedNodeTypes,
					env.llms,
					args.timeoutMs,
					env.nodeDefinitionDirs,
				)
			: createWorkflowGenerator(env.parsedNodeTypes, env.llms, args.featureFlags);

	// Create evaluators based on suite type
	const evaluators = createEvaluators({
		suite: args.suite,
		judgeLlm: env.llms.judge,
		parsedNodeTypes: env.parsedNodeTypes,

View on GitHub (pinned to 5ac6606e81)

Solutions

Export the key in the same shell that runs the eval: `export LANGSMITH_API_KEY=lsv2_sk_...`.
If using a `.env` loader, confirm it runs before `runV2Evaluation()`; otherwise prefix the command inline: `LANGSMITH_API_KEY=lsv2_sk_... pnpm eval --backend langsmith --dataset NAME`.
In CI, add `LANGSMITH_API_KEY` to the repository/worker secret store and verify the job actually receives it (print its length, never the value).

Example fix

// before
$ pnpm eval --backend langsmith --dataset my-dataset
Error: LangSmith client not initialized - check LANGSMITH_API_KEY
// after
$ export LANGSMITH_API_KEY=lsv2_sk_...
$ pnpm eval --backend langsmith --dataset my-dataset

Defensive patterns

Strategy: validation

Validate before calling

function assertLangsmithEnv(): void {
  const key = process.env.LANGSMITH_API_KEY;
  if (!key || key.trim() === '') {
    throw new Error('LANGSMITH_API_KEY is missing or empty; set it before using --backend langsmith');
  }
}
// call before runV2Evaluation when backend === 'langsmith'
if (backend === 'langsmith') assertLangsmithEnv();

Prevention

Store the key in your team secret manager and inject it into the eval job's env explicitly.
In CI, fail the job early if `LANGSMITH_API_KEY` is unset instead of letting it reach the CLI.
Print `process.env.LANGSMITH_API_KEY?.length` (never the value) at job start to confirm the secret is present.

When it happens

Trigger: `args.backend === 'langsmith'` AND `env.lsClient` is falsy after environment setup. Typically `LANGSMITH_API_KEY` is unset, set to an empty string, or only exported in a different shell than the one running the eval.

Common situations: Running evals in a fresh terminal, in CI without the secret in the matrix env, after switching accounts/projects, or when the key is stored in a `.env` file that the CLI does not auto-load.

Related errors

AI-assisted analysis of n8n-io/n8n@5ac6606e81 (2026-08-12). Data as JSON: /api/errors/a184a2644cde0f5a. Report an issue: GitHub.