Pulse Pick error reference

Every error this API returns carries a code. Build your integration against the code and the HTTP status. Do not match on message: it is a human-readable sentence and its wording can change between releases.

{
  "message": "Response not found.",
  "code": "pack.response.not_found"
}

Where no code applies the response omits the key entirely rather than sending "code": null, so an integration that reads only message keeps working.

This page is generated from the error catalogue published on the deployed OpenAPI spec, so it cannot drift from what the API actually returns. 28 codes are listed.

400 Bad Request

Retryable? No. Fix the request first.

Code	Message	Parameters
common.invalid_date_format	Invalid date format for {name}. Use ISO format 'YYYY-MM-DDTHH:MM:SS'.	name
history.invalid_cursor	Invalid cursor.	none
integration.api_config_not_found	Could not resolve API configuration '{config_id}'.	config_id
kpi.invalid_frequency	Invalid frequency. Use 'day', 'week', 'month' or 'hour'.	none
optimize.invalid_request_id	Invalid request id.	none
request.invalid_gzip	We couldn't decompress the request body. Make sure it's valid gzip content and try again.	none
warehouse.version_not_ready	Warehouse version is not ready. Please process it first.	none

401 Unauthorized

Retryable? No. Check your API key.

Code	Message	Parameters
auth.not_authenticated	Not authenticated.	none

402 Payment Required

Retryable? No. The subscription does not cover this action. Contact us to change your plan.

Code	Message	Parameters
subscription.not_subscribed	Your subscription does not cover this action.	none

404 Not Found

Retryable? No. The resource does not exist.

Code	Message	Parameters
optimize.postprocessing_step_not_found	Post-processing step not found with function_id '{function_id}'.	function_id
optimize.preprocessing_step_not_found	Pre-processing step not found with function_id '{function_id}'.	function_id
optimize.request_not_found	No optimization request found for id '{request_id}'.	request_id
warehouse.no_production_version	No production version found for warehouse '{warehouse_name}'. Mark a version as production first.	warehouse_name
warehouse.not_found_by_name	No warehouse found with name '{warehouse_name}'. Check the warehouse_name matches exactly.	warehouse_name
warehouse.version_not_found_for_site	Warehouse version not found for site '{site_name}'.	site_name

422 Unprocessable Content

Retryable? No. Fix the request first.

Code	Message	Parameters
request.invalid_body	The request body is invalid.	none
validation.bool_type	Field '{field}' must be true or false.	field
validation.duplicate_order_pick_ids	Provided (order ID, pick ID) combinations are not unique.	none
validation.duplicate_pick_ids	Provided pick_id's are not unique.	none
validation.float_type	Field '{field}' must be a number.	field
validation.int_type	Field '{field}' must be a whole number.	field
validation.invalid	Field '{field}' is invalid ({type}).	field, type
validation.missing	Field '{field}' is required.	field
validation.string_type	Field '{field}' must be text.	field
validation.too_short	Field '{field}' must not be empty.	field

429 Too Many Requests

Retryable? Yes, after a pause. This is a usage limit.

Code	Message	Parameters
algo.task_limit_exceeded	You have reached the limit of {limit} concurrent {algo_task_type} tasks ({current} running). Please wait for existing tasks to finish.	algo_task_type, current, limit

500 Internal Server Error

Retryable? Yes, once, with backoff. Contact support if it persists.

Code	Message	Parameters
optimize.failed	Optimization task failed.	none

202 Accepted, still processing

Retryable? Yes. The result is not ready yet; poll again or use a webhook.

Code	Message	Parameters
optimize.still_processing	Optimization still in progress.	none

Errors that do not come from the API

These responses are produced by the infrastructure in front of the API, before a request reaches the application or after it has already returned. They never carry a code, and their body is not in the shape described above. Do not assume code is always present.

Status	Usual cause	What to do
413	The request body exceeded the platform limit before reaching us.	Send fewer items per request, or compress the body.
502	The response was too large to return, or the function failed outright.	Split the work across smaller requests. If it persists on a small request, contact support.
504	The request took longer than the gateway allows (currently 120 seconds).	Use the asynchronous flow rather than waiting on a synchronous response.
429	Platform-level throttling, distinct from the account limits listed above.	Back off and retry. If it recurs at normal volume, contact support.

A 502 after a request that clearly succeeded usually means the result was too large to return in one response. That is the case the asynchronous flow exists for: submit the work, then fetch the result separately.

Validation errors

A malformed request body is rejected before it reaches an endpoint and returns a different shape: a detail list with one entry per problem. Each entry names the offending field in loc and the failure kind in type.

{
  "detail": [
    {"loc": ["body", "orders"], "msg": "Input should be a valid list", "type": "list_type"}
  ]
}