v0.0.4

What's new

• Extract endpoint — POST /v1/extract parses a document and returns structured JSON via OpenRouter (default google/gemini-2.5-flash)
• Async extract endpoint — POST /v1/extract/async enqueues an extract job and returns a job_id
• JSON Schema support — pass a schema field to constrain extract output via structured outputs
• Extract queue — wired into BullMQ, Bull Board UI, and GET /v1/job/:job_id lookup
• UUID job IDs — parse and extract jobs now use UUIDs

v0.0.3

What's new

• Async parse endpoint — POST /v1/parse/async returns a job_id immediately (requires REDIS_URL)
• Job status endpoint — GET /v1/job/:job_id to poll for result
• BullMQ integration — parse jobs processed via Redis-backed queue with worker
• Bull Board UI — available at /v1/queues when Redis is configured
• Docker fixes — correct Redis/S3 networking in docker-compose, Bull Board assets included in image

v0.0.2

Optional S3 document staging for parse inputs.

Container image: ghcr.io/ocrbase-hq/ocrbase:0.0.2
Digest: sha256:3daf56805eb799f1b0c258b980b2fcffc63560640c4c0363a58408341a5585d9

v0.0.1

Full Changelog: https://github.com/ocrbase-hq/ocrbase/commits/v0.0.1