Before you connect: Connecting Dextr to ChatGPT or Claude makes contact information and other Dextr library content available to those services through the access you authorize. Consider privacy obligations and your company’s policies before connecting, and review the AI provider’s data controls.
Apple’s built-in iPhone Contacts app handles the basics: saving a name, finding a phone number, and sending an email. But as your address book grows, it can feel too basic for the way you actually know people. A long list of contact cards gives you little help deciding who to follow up with, remembering where you met, or organizing a network of friends, clients, and colleagues.
Dextr is a contacts app and personal CRM for iPhone, iPad, and Mac. A personal CRM helps you manage the context around your relationships as well as the contact details. You can make Dextr the app you open every day to find people, start calls and messages, organize your contacts, and plan your next conversation.
That makes Dextr a more capable alternative to the native iPhone Contacts app when you want to do more with your address book:
- Tags and groups organize people by interests, projects, events, or professional roles.
- Journal entries, First Met details, and introductions help you remember your shared history.
- Relationships and the Social Graph show how people are connected to you and to one another.
- Custom fields and reminders turn details such as a client’s project stage or your next follow-up into information you can act on.
You can bring your existing contacts with you. Install Dextr on your iPhone, open Home → Settings → Contacts Sync, and turn on Enable Contacts Sync. Follow the setup prompts, choose the contact access you want to allow, and identify your own contact card. Let the import finish, then use Sync Now and Last Sync Results to check that your contacts came through.
Contacts Sync brings the Apple Contacts records you allow into Dextr and sends supported contact-detail changes back through one Primary Sync Device, usually your everyday iPhone. That lets you organize people in Dextr while keeping supported names, numbers, and email addresses available to the iPhone apps that use Apple Contacts.
Dextr’s extra context—including tags, custom fields, journal entries, and relationship connections—belongs to Dextr rather than becoming equivalent fields in Apple Contacts. Dextr’s separate iCloud Sync carries its own data between your devices. For the complete setup and sync details, follow the import and sync guide.
Once your contacts are in Dextr, AI can help you organize them at a larger scale. Dextr’s MCP connection lets ChatGPT or Claude work with your contact records: find matching people, propose tags, populate custom fields, build a follow-up plan, or prepare a report. MCP stands for Model Context Protocol; it gives the assistant tools for working with Dextr.
The examples below work as starting prompts for ChatGPT or Claude with Dextr connected. Dextr Pro is required, and the Dextr Mac app must be installed and open on the same Mac while the connection is in use. Available actions depend on the Dextr version and tools enabled in your conversation.
The dark-mode screenshots show Dextr’s Mac setup dialogs and iPad app views with fictional demonstration data. They illustrate the connection and the records you can organize; they are not captures of the AI executing these prompts.
Connect Dextr and Start with a Preview
Begin with the Dextr setup guide for ChatGPT and Claude. In the Mac app, choose Dextr → Install Dextr in ChatGPT or Dextr → Install Dextr in Claude, then finish installation in the selected AI app.
You can jump directly to the ChatGPT installation steps or Claude installation steps.
Start with a lookup before asking for changes:
Use-case example: Use Dextr to find Alex. Show the matching contacts, their organizations, job titles, tags, and relationship to me. Do not change anything. If more than one person matches, ask me which one I mean.
For larger tasks, use a simple sequence: find the records, preview the proposed changes, approve the specific changes, and verify the result. Ask the assistant to identify any unavailable action instead of claiming it completed a step its tools cannot perform.
Organize People with Tags and Clear Criteria
Tags make useful groups out of a long contact list. You might group people by a recorded job title, an event you met at, an interest mentioned in your notes, or a project you are working on together.
Choose a rule the assistant can explain:
Use-case example: Find all active Dextr contacts whose recorded job title contains “Designer.” Retrieve every result page and show their names, job titles, and current tags. Propose adding a Design tag, reusing an equivalent existing tag if one exists. Preserve every current tag. Wait for my approval, then apply only the approved changes and read the contacts back to verify them.
The same approach works for an event group or a shared interest. Ask the assistant to show the supporting event or note for each match, and keep uncertain cases in a separate review list. A missing job title is a reason to ask for more information, not a reason to guess.
For more ways to structure groups, see how to tag a contact in Dextr.
Import a Spreadsheet with a Report of Proposed Changes
A spreadsheet might contain new contacts, updates to people already in Dextr, and duplicate rows. AI can help map the columns and prepare the import, but the most useful first output is a report of what would change.
Attach the spreadsheet or CSV to your AI conversation and ask:
Use-case example: Compare this spreadsheet with my Dextr contacts. Before making changes, generate a proposed-changes report. For each row, show the matched contact or proposed new contact, current values, proposed values, possible duplicates, ambiguous matches, and unsupported columns. Map names, company, job title, email, phone, tags, and custom fields where supported. Include totals for additions, updates, and skipped rows. Do not write anything until I approve the report.
Review the report, resolve ambiguous matches, and approve a small batch first. Have the assistant preserve existing phone numbers, email addresses, tags, and other values you did not ask to replace.
Afterward, ask for a verification report showing what was added, updated, skipped, or left unfinished. Possible duplicates should remain review items unless you have explicitly chosen how to resolve them.
You can also use this workflow for a list copied from a document. The important part is giving the assistant a clear source and checking its proposed mapping before it writes to Dextr.
Create and Populate Custom Fields
Custom fields give a contact list structure beyond names and phone numbers. A consultant might track project stage and budget. A contractor might track Pipeline Stage: Warm Lead, Deal Value: $12,500, and Service: Artificial Turf Installation.
Creating a field definition and setting its value on a contact are separate actions. A capable Dextr MCP build supports both; ask the assistant to reuse existing definitions before creating new ones.
Use-case example: Check my existing Dextr custom fields. Propose any missing fields for Pipeline Stage as text, Deal Value as a number formatted in US dollars, and Service as text. For the lead I select, preview Pipeline Stage = Warm Lead, Deal Value = $12,500, and Service = Artificial Turf Installation. Show the field definitions and proposed contact changes before writing. After I approve, create only missing definitions, populate the approved values, preserve all other custom fields, and read the contact back to verify the result.
Keep field names and units consistent. A budget field is easier to compare when all amounts use the same currency. If your installed Dextr version does not expose a required write action, update Dextr or complete that step in the app.
See how to create and use custom fields for more personal and professional examples.
Record Relationships and Introductions
A tag describes a category. A relationship records how people are connected. Dextr can distinguish your relationship to a person from a relationship between two contacts, and it can record who introduced you.
Use confirmed facts and name the direction of a relationship:
Use-case example: Find Morgan and Taylor in Dextr and show their current relationships. I confirm that Morgan is Taylor’s manager. Preview the relationship type and direction before saving it. Separately, show the active contacts at Example Studio and propose setting their relationship to me to Colleague only for the people I confirm. Preserve unrelated relationships and wait for my approval before making changes.
For introductions, a request such as “Record that Amber introduced me to Alice at the book club” is more useful than “Connect these people.” Have the assistant confirm the exact contacts and any existing introducer information first.
A shared employer or a senior job title alone does not establish a reporting relationship. Ask the assistant to flag uncertainty and use the relationship guide when you need to review the result in Dextr.
Turn Conversation Notes into Useful Contact History
Short notes are often enough to preserve something you would otherwise forget: a person’s current project, a recommendation they shared, or what you promised to send. AI can organize those notes into a concise journal entry associated with the right contact.
Give the assistant the notes you want summarized, along with the date and type of interaction:
Use-case example: Turn the notes I pasted below into a concise Dextr journal entry for Sofia. Include the conversation date, how we spoke, the subjects discussed, and the follow-up I actually promised. Use only facts in my notes. Show the matched contact and draft entry first. After I approve, save one entry, check for an existing matching entry to avoid a duplicate, and read it back.
You can also ask the assistant to summarize history already recorded in Dextr before a meeting. Ask it to identify which entries support each point and distinguish saved facts from suggestions.
The MCP connection does not mean the assistant automatically has access to all your messages, recordings, or files. Supply the source material you want it to use, or enable an appropriate separate tool. Dextr’s attachment-list tools provide file metadata rather than the contents of every attached document.
Build a Follow-up Plan with Reminders
Organization should help you decide what to do next. A useful review might focus on a project, a tag, people you recently met, or contacts with no recorded interaction for a chosen period.
Have the assistant propose a manageable plan before creating reminders:
Use-case example: Review active contacts tagged Design. Show people whose last recorded interaction was more than 90 days ago, and list people with no recorded interaction separately. Check their existing open reminders. Propose five follow-ups for next week with a person, reason, reminder title, date, and time zone. Do not create duplicate reminders. Wait for me to approve the people and dates, then create only those reminders and verify them.
“No recorded interaction” does not necessarily mean you have not spoken. The assistant can only review the history available to its tools.
For a simple request, name the contact and next step directly: “Remind me to ask Ryan about his project next Tuesday at 10 a.m. in my time zone.” Confirm the resolved calendar date and time zone before saving. Creating a reminder or recording an interaction does not itself send a message.
Generate Event Reports and Spreadsheet Exports
An event gives a report a useful scope. Instead of asking for a summary of your entire address book, ask for the people associated with one meetup, the information you have recorded about them, and a clear list of missing details.
Try a report you can review or move into a spreadsheet:
Use-case example: Find the Coffee & Tech Meetup event I select in Dextr. Retrieve its associated people and their available contact details, tags, relationship to me, and recorded follow-up notes. Produce a report grouped by organization, plus a spreadsheet-ready table with name, company, job title, email, tags, and next action. Include the event date, total people reviewed, missing data, and any incomplete results. Make no changes to Dextr.
The assistant can format retrieved information as a table or CSV content. Downloadable files depend on the file tools available in ChatGPT or Claude. Ask it to use only the columns you need and check the row count before sharing the output. A selected export is not a complete Dextr backup.
You can use the same pattern for a sales pipeline review, a list of contacts missing job titles, or a weekly follow-up report. When filtering a custom text field such as Pipeline Stage, ask the assistant to read and compare the actual stored values if a direct text comparison is unavailable.
Start with one useful change: a better tag, a reviewed spreadsheet import, or a short follow-up list. Then build a repeatable routine around the results. Dextr holds the people and context; the AI helps you review that information, propose organization, and carry out the changes you approve.
Connect Dextr to ChatGPT or Claude to try your first read-only lookup, then choose one example above for your next task.