How to Use Claude With Your LinkedIn Network: Setup, 18 Prompts, and the Limits

Your LinkedIn network is probably the most useful list of people you own, and the hardest one to search. You remember that someone you met years ago went into climate tech, but not their name. You know a few investors liked your posts last month, but the notifications are long gone.

Claude is good at exactly this kind of fuzzy question. It just needs access to the list. This guide covers how to give it that access through wait who?, what to ask once it's connected, and where the edges are.

What a connector is, in plain words

On its own, Claude only knows what is in the conversation and what it learned in training. It has no idea who is in your LinkedIn network. A connector is a bridge to another app. Anthropic's help center puts it like this: "Connectors let Claude access your apps and services, retrieve your data, and take actions within connected services."^[3]

Under the hood, connectors use the Model Context Protocol, or MCP. The project's own docs call it "an open-source standard for connecting AI applications to external systems," and compare it to a USB-C port: one standard plug instead of a different cable for every device.^[1] Anthropic open-sourced MCP in November 2024, describing the problem it solves as AI models being "trapped behind information silos," where "every new data source requires its own custom implementation."^[2]

In practice, an app that supports MCP offers Claude a short list of tools, each with a name and a description. When you ask a question, Claude decides which tool to call, the app returns the data, and Claude writes the answer from what came back.

How to connect wait who? to Claude

wait who? is listed in Anthropic's connector directory, so you add it with one click. Open claude.ai/directory/waitwho, or search "wait who" under Customize > Connectors in Claude.^[5]

Before you start

• A wait who? account with LinkedIn connected and your network synced.
• An active membership or free week.
• Saving drafts from Claude needs a plan that includes Content Studio. Everything else is read-only and works without it.

On a Free, Pro or Max plan

1. Open the wait who? listing in Claude's connector directory and click Connect.^[5]
2. Sign in with your wait who? account when asked, and press Continue to approve access.
3. Using another MCP client, or Claude without the directory? Add a custom connector with the URL https://waitwho.ai/mcp.^[4]

On a Team or Enterprise plan

An Owner enables wait who? once under Organization settings > Connectors. After that, each member finds it under Customize > Connectors and clicks Connect to sign in with their own account.^[5] Each person only ever sees their own network.

Turn it on in a chat

Click the + button in the chat box, open Connectors, and switch wait who? on for that conversation.^[5] Then ask a question. Claude may ask you to approve a tool before it uses it, and Anthropic suggests reading those requests carefully.^[4]

One honest note. Being listed in the directory means Anthropic reviewed the listing, not that it audited the service, and you should only connect services you trust. The rest of this guide lists exactly what ours can and can't do, so you can decide.

What Claude can do with wait who?

Claude gets eight tools. Five only read. Three write, and only to your own account: drafts in your post calendar and private notes about people in your network.

• Today. Your daily picture: people worth reaching out to for your goal, posts scheduled this week, and your recent reach.
• Find people. Searches your own network by topic, industry, role, seniority or company, best fit first. It can limit results to people who engaged with your posts.
• One person's context. Role, company, work history, when you connected, whether you've messaged them, and how they engaged with your posts, including what they wrote in comments.
• Post vs comment reach. Views on your posts next to views on your comments, for up to the last 90 days, plus your top comments.
• List drafts and scheduled posts. What's in your calendar.
• Save a draft. Saves a post as a draft, with an optional planned date. Drafts are never posted.
• Edit a draft. Changes the text or planned date of a draft.

Claude can chain these. "Find fintech founders who engaged with my posts, then tell me more about the top three" is one search plus three lookups, and you get one answer. The full reference is on our Claude connector page.

18 prompts, grouped by what you're working on

Swap in your own names and fields. The more specific the question, the better the answer, because Claude turns your words into search terms.

Finding customers

• "Who in my network is a head of operations or COO at a logistics company? Best fit first."
• "Which people who engaged with my posts work in e-commerce? I'm looking for early customers."
• "Find people at [company name]. Who do I know there, and have I ever messaged any of them?"

Pair these with the approach in our guide to finding your first customers in your network.

Raising a round

• "Find investors, angels and VCs in my network. Which of them have engaged with my posts?"
• "Tell me about [investor name] before our call. How long have we been connected and what have they reacted to?"
• "Draft a short post announcing we passed 100 customers, in a plain tone, and save it as a draft for Tuesday at 9am my time."

Hiring

• "Who do I know with 'recruiter' or 'talent' in their title? I want to ask for referrals for a senior designer role."
• "Find senior engineers in my network who work at fintech companies."
• "Look up [name]. What's their work history, and when did they start their current role?"

Partnerships

• "Who do I know at agencies that work with Shopify brands?"
• "Find heads of partnerships or business development in my network, and tell me which ones have engaged with my posts."

Your brand and content

• "Compare my comment views and post views over the last 90 days. What should I do more of?"
• "What's in my content calendar for the next two weeks? Where are the gaps?"
• "Turn these meeting notes into a LinkedIn post and save it as a draft for Thursday morning." (paste your notes)
• "Make my draft about hiring shorter and move it to Friday at 8:30."

Keeping in touch

• "What should I do on LinkedIn today?"
• "Who has engaged with my posts more than once but never heard from me? Suggest three people to thank."
• "Who did I connect with recently? Give me a one-line reason to say hello to each."

For the message itself, Claude can draft it in the chat. You copy it, change what doesn't sound like you, and send it on LinkedIn yourself. Our post on reconnecting without the cringe has a three-line format that works well as a starting point.

What Claude can't do

Some of these are deliberate choices, and some are just how the data works.

• It can't act on LinkedIn. No posting, scheduling, commenting, reacting, connection requests or messages. The connector never calls LinkedIn at all. It reads data wait who? already synced, and the only thing it writes is a draft in your calendar. You review and schedule drafts yourself in wait who?.
• It only knows your own network. Your synced connections and followers. It can't search LinkedIn at large or see other people's connections.
• It's as fresh as your last sync. If someone changed jobs yesterday, Claude won't know until wait who? syncs again.
• It searches what's on the profile. Searches match names, headlines, companies and titles, plus industry, role and seniority. If someone's headline says "Building something new," a search for "fintech" may miss them.
• It can get things wrong. Claude summarizes what the tools return, and summaries can be off. Check the details before you mention them to someone, especially dates and job titles.

If you want a tool to never be used, Anthropic's docs explain how to set a single tool's permission to "Blocked" under Customize > Connectors, and recommend only clicking "Always allow" for servers you trust.^[5] For example, you could block the two draft tools and keep the connector read-only.

Privacy: who sees what

• Claude only sees your wait who? data after you connect and approve it. You can turn the connector off for any conversation, or remove it entirely.^[5]
• We receive the tool requests, not your conversation. When Claude calls a tool, we get the request, like a search term or the text of a draft you asked it to save. We don't receive or store your chats with Claude. We log the tool name, the time and whether it worked, never the content.
• The assistant provider processes what you share in the chat. That includes the answers our tools return. For Claude, that's Anthropic, under its own terms.
• No credits. Reading your data and saving drafts are free. Publishing a post you schedule in wait who? uses one credit, as it always does.

The full details are in the AI assistants section of our privacy policy.

A good first five minutes

1. Ask "What should I do on LinkedIn today?" to see that it's working and get a starting list.
2. Ask one search question for whatever you're working on this month.
3. Pick one person from the results and ask for their full context.
4. Ask Claude to draft a short, specific message to that person. Edit it and send it yourself.

If you have a call coming up, our five-minute meeting prep guide has a checklist and prompts built for exactly that.

FAQ

Sources

1. What is the Model Context Protocol (MCP)?. Model Context Protocol documentation, modelcontextprotocol.io.
2. Introducing the Model Context Protocol. Anthropic, November 25, 2024.
3. Use connectors to extend Claude's capabilities. Claude Help Center.
4. Get started with custom connectors using remote MCP. Claude Help Center.
5. Add a connector that isn't in the directory. Claude documentation, claude.com.