CatchAll API changelog

​

Job modes, plan limits, and response improvements

• Job modes: Submit jobs in lite or base mode with the new mode parameter. Lite mode provides lightweight extraction with titles and citations only. Base mode runs the full pipeline with validation and enrichment.
• Plan limits: Retrieve plan features and current usage, including concurrency slots, max results, lookback limits, and monthly credits.

• POST /catchAll/user/limits — Get plan features and current usage.

• GET /catchAll/pull/{job_id} now returns mode and full validator/enrichment objects instead of name-only arrays.
• GET /catchAll/jobs/user now returns mode and masked user_key for each job.
• GET /catchAll/monitors now returns masked user_key for each monitor.
• Fixed duplicate confidence values appearing in enrichment records.
• Fixed pagination handling in the pull endpoint.
• Improved multi-word query handling for lite mode.
• Gateway authentication caching reduces response times.

​

Monitor controls and response improvements

• Monitor limit: Control the maximum number of records per monitor run with the new limit parameter.
• Monitor backfill: Choose whether to fill the data gap before the first scheduled run with the new backfill parameter.

• GET /catchAll/monitors now supports page and page_size parameters and sorts by creation date.
• GET /catchAll/pull/{job_id} now returns the error field for failed jobs and the limit applied to the job.
• POST /catchAll/continue now correctly enforces result limits. new_limit defaults to your plan maximum if omitted.
• GET /catchAll/jobs/user now returns a structured response object with pagination metadata instead of a flat array.
• Consistent status representation across the status and pull endpoints.
• Improved caching and overall response performance.
• Submitting a validator or enrichment with an empty name now returns a validation error.

​

Stable release with enhanced control and structured company enrichments

• Initialize endpoint: Get suggested validators, enrichments, and date ranges before submitting jobs. Returns date_modification_message when requested dates exceed plan limits.
• Date range controls: Specify start_date and end_date in submit requests. System validates against plan limits and returns 400 errors for invalid ranges.
• Custom validators and enrichments: Define custom validation criteria and extraction fields directly in submit requests instead of relying on automatic generation.
• Company enrichment type: Extract structured company data including name, alternative names, website candidates, people, and address using the new company enrichment type.
• Paginated user jobs: List user jobs endpoint now supports page and page_size parameters for efficient retrieval of job history.

• POST /catchAll/initialize — Get validator, enrichment, and date range suggestions.
• Enhanced POST /catchAll/submit — Now accepts date ranges, validators, and enrichments.
• Enhanced GET /catchAll/jobs/user — Now supports pagination.

• Enhanced deduplication logic reduces duplicate records across job executions.
• Status endpoint returns more accurate current processing stage.
• User jobs endpoint returns results sorted by creation date (most recent first).

​

Job continuation and early result access

• Job continuation: Submit jobs with the optional limit parameter for fast testing. Review results and continue jobs with /catchAll/continue to process more records without restarting.
• Early result access: Pull results during the enriching stage without waiting for job completion. A new progress_validated field shows validation progress as results become available in batches.
• Monitor updates: Update webhook configuration for existing monitors without recreating them.

• POST /catchAll/continue — Continue completed job with higher limit.
• PATCH /catchAll/monitors/{monitor_id} — Update monitor webhook configuration.

• Expanded search range provides broader coverage for improved recall.

​

API response enhancements and monitor configuration

• Enrichments metadata: Job pull endpoint returns enrichments array listing all extracted field names.
• Human-readable schedules: Monitor list includes schedule_human_readable alongside cron expressions.
• Webhook visibility: Monitor list exposes full webhook configuration including URL, method, and headers.

• Pull endpoints return all citations without limits.
• Monitors require minimum 24-hour intervals between executions.
• First webhook execution includes complete reference job results.
• Improved validation error messages for invalid schedules.

​

Status tracking and monitor improvements

• Progress tracking: Status endpoint now returns detailed steps array showing completion state for each processing stage.
• Monitor timestamps: Monitor records include added_on and updated_on fields to track when each data record was first collected and last updated across job executions.
• Time-constrained jobs: Monitors now work with reference jobs that include time constraints. The system automatically adapts date-specific validators for each execution.

• GET /catchAll/jobs/user now returns only user-created jobs, excluding repeated monitor jobs.
• Monitor responses use run_info structure with first_run and last_run timestamps instead of date_range for clearer execution metadata.
• Improved error handling in monitor creation endpoint with specific validation messages for invalid schedules and incomplete reference jobs.

​

Monitor webhooks and deduplication

• Webhook notifications: Receive real-time POST notifications when scheduled monitor jobs complete, including monitor ID, latest job ID, and record counts.
• Intelligent deduplication: Monitors automatically eliminate duplicate records across job executions, sending only new results to webhooks.
• Monitor management: Enable or disable monitors to control automated job execution.

• GET /catchAll/monitors — List all monitors for your API key.
• POST /catchAll/monitors/{monitor_id}/enable — Enable a disabled monitor.
• POST /catchAll/monitors/{monitor_id}/disable — Disable an active monitor.

• Reduced default date range from 14 days to 5 days for faster job processing.
• Limited maximum date range to 30 days to maintain optimal performance.
• Fixed special characters displaying incorrectly in web page content.
• Improved error handling when monitor jobs have incomplete data.

​

Pagination and reliability improvements

• GET /catchAll/jobs/user — List all jobs created with your API key.
• GET /catchAll/monitors/pull/{monitor_id} — Retrieve aggregated results from all monitor jobs.

• Fixed pagination returning all records instead of the requested page.
• Enhanced pagination support for handling large result sets with page and page_size parameters.
• Fixed jobs failing to complete during data extraction.
• Fixed timezone errors when scheduling monitor jobs.
• Fixed monitors failing to execute all configured queries.

​

Monitors feature

• Schedule automation: Create monitors that run queries on schedules defined in plain text format (e.g., “every day at 12 PM UTC”).
• Job tracking: Track all jobs associated with a monitor.
• Execution history: Retrieve job execution history and status for each monitor.
• Webhook notifications: Get notified when scheduled jobs complete.

• POST /catchAll/monitors/create — Create scheduled monitor.
• GET /catchAll/monitors/{monitor_id}/jobs — List monitor jobs.

• Added date range information to results showing the time period covered.
• Fixed incorrect date filtering in search results.

​

Initial beta release

• Plain text input: Submit queries in plain text and receive structured event data with source citations.
• Dynamic schemas: Response structures adapt to your query, generating relevant fields automatically.
• Real-time status tracking: Track job progress from submission through completion, including analysis, retrieval, clustering, validation, and extraction.

• POST /catchAll/submit — Create jobs using a plain text queries.
• GET /catchAll/status/{job_id} — Check a job processing status.
• GET /catchAll/pull/{job_id} — Retrieve structured results with citations.