Normalize countries
Country values are enough. Names, emails and phone numbers stay hidden.
Data Gate
Choose what your agent can read and change.
Restrict visible columns, permitted actions and writable fields in the ListPlus connection. The permissions belong to the connection, not to the prompt.
Visible columns
Hide fields the task does not need. Restrict the connection to selected sublists.
Permitted changes
Allow specific actions and writable columns. Fill only empty fields where that is the task.
Connection limits
Expiry date, requests per minute and row caps, per connection.
Task examples
Which fields does the task need?
Configure data access separately from the instruction you give your agent.
Find job titles
Name and company give the context. Allow updates to empty title fields only.
Verify emails
The check needs the email address. Other contact fields stay hidden.
Connection settings
Set read access and write permissions.
Every setting below exists in the connection dialog. The prompt cannot widen them.
- Visible columns and sublists
- Choose which fields and which records the connection exposes. Hidden columns never leave ListPlus.
- Writable columns
- Allow changes to selected fields only. Everything else is read-only for the agent.
- Only fill empty
- Missing values get filled, populated fields stay as they are.
- Allowed actions
- Enable just the checks, enrichments and research tools the task needs. Presets: Read Only, Data Assistant, Enrichment, Data Collector, Full Access, Custom.
- Limits
- Expiry date, rate limit per minute and a maximum of records per enrichment request.
Country normalization
Normalize repeated values once.
The agent works with distinct values and applies the mapping to the whole list.
Values in the list
DE · Germany · DE · Deutschland · Germany · DE
Distinct values
DE · Germany · Deutschland
Mapping
Each label maps to Germany. Three decisions instead of one per row.
Update
The mapping is applied to the country field only, with dry run first if you want.
Connection details
Explore agent workflows Connect your agent.
MCP and REST access, setup steps and plan requirements are on the AI agents page.
Good to know.
Is this the same as telling the AI to ignore a column?
No. An instruction is not a boundary. Configure what the connection actually exposes; hidden columns never reach the agent.
Can an agent ever see contact information?
Yes, when the task needs it. Person research needs a name and company, email verification needs the address. Choose the scope per task.
Can I protect existing values?
Yes. Restrict writable columns and switch on Only fill empty. Reads and writes are separate settings.
Can a connection be temporary?
Yes. Set an expiry date, a rate limit per minute and a maximum of records per enrichment request. You can also deactivate a connection at any time.
Continue exploring
Start with anything. End with contacts that are ready.
Try the sample list or bring your own starting point.