ax-batch-analysis

Batch AI analysis on internal data can be automated using Claude Code subscriptions instead of API credits. This Skill processes items from a remote D1 database in batches to handle tasks like clustering and evaluation without extra costs.

Batch data processing supports radar item clustering, idea generation, ontology entity and relation extraction, and item quality evaluation. You can execute these tasks using radar, ontology, eval, or all modes.

Yes, batch AI analysis connects directly to a remote D1 database to retrieve and process items. It extracts ontology entities and relations from your stored data using Claude Code subscriptions.

Batch AI analysis handles up to 5 items per batch and includes error skipping for individual item failures. This ensures the batch job continues processing remaining items even if one fails.

Running batch AI analysis in eval mode evaluates item quality and serves as a supplement to production cron jobs. This approach corrects omissions and handles bulk processing efficiently overnight.

Batch size is limited to 5 items per batch when running ontology extraction or other AI analysis modes on a D1 database. This constraint optimizes processing stability and manages subscription usage.