databar-bulk-enrichment

Databar Bulk Enrichment

Enrich a list of records in a single operation and return results inline. Unlike the table-driven skill, this returns data directly to the conversation without creating a persistent table.

When to use this vs table-driven enrichment

• Bulk enrichment (this skill): User wants quick results returned directly. Good for smaller lists (up to 100 items).
• Table-driven enrichment (databar-table-enrichment skill): User wants a persistent table with a link, or has more than 100 items, or wants to run multiple enrichments on the same dataset.

Workflow

1. Parse the user's input list. Accept any format: CSV, JSON, comma-separated, numbered list, plain text. Extract each record's fields.

2. Determine enrichment vs waterfall.

   • If the user wants to try multiple providers (e.g. "find emails using all available sources"), use search_waterfalls and run_bulk_waterfall. See the databar-waterfall skill.
   • Otherwise, use search_enrichments and run_bulk_enrichment.

3. Find the right enrichment. Call search_enrichments with a query matching the task.

4. Inspect it. Call get_enrichment_details to check required parameters, price per record, and choices for any select/mselect params.

5. Resolve choices for select/mselect params. If any param has a choices object, you must use valid id values (not display names) when building params_list.

   • choices.mode = "inline" — pick from choices.items[].id directly.
   • choices.mode = "remote" — call get_param_choices({ enrichment_id, param_name }) to get valid options. Use q to search if needed.
   • mselect params accept an array of ids.

6. Validate the list.

   • Max 100 items per request. If the user provides more, warn them and either:
     • Truncate to 100 with their approval
     • Suggest using the table-driven skill for larger datasets
   • Ensure each record has the required parameters. Flag any incomplete records.