Overview
The Error Ingestion API receives error events from your application via the SDK. It supports both single error ingestion and batch uploads for efficient data transmission.Authentication
Error ingestion uses API key in payload authentication. Include your project API key in the request body. See Authentication for details.Single Error Ingestion
Send a single error event to Proliferate.Endpoint
Request Schema
Your project API key (format:
pk_...)Unique identifier for this event (generated by SDK)
When the error occurred
Event type:
exception or messageSeverity level:
error, warning, or infoEnvironment where error occurred (e.g.,
production, staging)Release version or git commit SHA
URL where the error occurred
Browser or client user agent string
Exception details (required if
type is exception)Custom message (required if
type is message)User context
Account/organization context (B2B apps)
Session identifier for replay correlation
Window identifier (for multi-tab tracking)
List of events leading up to the error
Additional custom context data
Response
Success (202 Accepted):Batch Error Ingestion
Send multiple error events in a single request for efficiency. The SDK uses this endpoint to batch errors collected over a time window.Endpoint
Request Schema
Body: Array of error events (max 100 events per batch) Each event follows the same schema as single error ingestion.Response
Success (202 Accepted):Error Grouping
Errors are automatically grouped into issues based on a fingerprint derived from:- Exception type
- Top 3 stack frames (function name + file name)
What Happens After Ingestion?
1
Fingerprinting
Error is fingerprinted based on exception type and stack trace
2
Issue Creation or Update
If an issue with this fingerprint exists, increment its count. Otherwise, create a new issue.
3
User & Account Tracking
If user/account context is provided, track affected users and accounts
4
Stack Trace Resolution
If
release is provided and source maps exist, resolve minified stack traces to original source5
Notifications
Trigger notifications based on organization preferences
Best Practices
Always Include Context
The more context you provide, the easier it is to debug:- User context: Identify which users are affected
- Account context: See which customers are impacted (B2B apps)
- Release: Enable source map resolution
- Breadcrumbs: Understand the sequence of events
- Extra data: Add custom context relevant to your app
Use Consistent IDs
- User IDs: Use the same ID format across your app
- Account IDs: Use your internal account/organization ID
- Event IDs: Let the SDK generate these (UUIDs)
Set Appropriate Levels
- error: Unhandled exceptions, critical failures
- warning: Handled errors, recoverable issues
- info: Informational messages (use sparingly)
Batch When Possible
The SDK automatically batches errors to reduce network overhead. If implementing custom ingestion, batch errors collected within a few seconds.Handle Rate Limits
If you hit rate limits (unlikely with normal usage):- Implement exponential backoff
- Queue events locally and retry
- Alert your team to investigate the error spike

