build-list

Instructions

Setup state. Not yet configured for this org. Load the Readme sub-page once to capture the org's default column plan, sourcing tier order, and activation defaults. (Rewrite this paragraph via swan-update-skill after setup so future runs read the current config and skip re-checking.)

The mental model — column ladder

Every list build is a table. Each column has four properties:

1. Input — which earlier columns it reads.
2. Cost tier — free filter / free elimination / cheap enrichment / AI research / paid contact data.
3. Run-when condition — which upstream Boolean gates it (almost always icp_pass).
4. Fallback — what to write when it can't resolve.

You never "search and figure out what to do with the results." You write the column plan, show it to the user, then run it. Cheap filters kill the most rows; only survivors reach the expensive columns.