Maintaining Orgs

Keep indexed organizations current by fetching their sources and regenerating overviews. Driven from Claude Code using the installed releases CLI against the production API. Two pieces compose:

1. Fetch — pull every active source so the database reflects what's been published. CLI does the work; no AI required.
2. Regenerate overview — re-derive the org's AI summary from the freshened release set. AI work; runs through the regenerating-overviews skill (which carries the prompt + workflow).

There is intentionally no single-command "refresh in one step" today. Compose the two pieces below.

When to Use

• Periodic maintenance: "update the top 10 orgs"
• Before analysis: ensure data is fresh before competitive intel or trend work
• After onboarding: first full fetch + overview for a newly added org
• Spot checks: "is Stripe up to date?"

On-demand orgs (discovery = 'on_demand') are hidden stubs materialized by the on-demand lookup endpoint. They get embeddings and cron fetches but no overviews or summarization. The releases admin overview plan / overview list endpoints already filter to curated orgs server-side via the organizations_public view (which excludes on_demand), so manifest-driven sweeps don't need a separate filter pass. To promote an on-demand org to curated, use releases admin org update <slug> --discovery curated.

Single Org Update