primitives job store

Job store

Active contributors: groktopus

Purpose

The job store provides Valkey-backed CRUD operations for all async job types (agent, crawl, extract, batch scrape, llmstxt). It is used by the API to create jobs and by workers to update status.

Key schema

Jobs are stored across two Valkey keys:

job:{id}:meta -- JSON with status, kind, created_at, completed_at, error
job:{id}:data -- JSON with result data (set on completion)

Both keys expire after 24 hours via Valkey TTL.

Methods

Method	Description
`create_job(kind, payload)`	Creates a new job with UUID, returns the ID
`get_job(job_id)`	Returns job metadata with attached data
`complete_job(job_id, data)`	Sets status to "completed", stores result
`fail_job(job_id, error)`	Sets status to "failed", stores error
`cancel_job(job_id)`	Cancels a processing job, returns False if not found or already done
`list_active_jobs(kind, status, limit)`	Scans Valkey for jobs matching status and optional kind filter