The safest LinkedIn CLI for your agents.
tth connects Claude Code, Codex, Cursor or any agent with a terminal to your own LinkedIn account. No MCP tool definitions loaded into every session: your agent reads --help when it needs a command.
Your agent reads your LinkedIn conversations, searches your network, enriches profiles with email and phone, and drafts the messages. You review them and hit send.
LinkedIn has no official CLI, and its public API doesn't give access to your private messages. Every LinkedIn CLI is third-party, and most fall into one of three groups:
| Open-source scrapers | Cloud LinkedIn tools | TalkToHumans | |
|---|---|---|---|
| Pace | Whatever pace your agent sets | Daily caps sized for campaigns | Human pace, enforced: daily limits plus a gap between every action, per LinkedIn account. Nobody can override them |
| What 100 agent calls do | 100 trips to LinkedIn | 100 trips to LinkedIn, from their servers | Read a synced cache, so no more LinkedIn activity than one call |
| Where LinkedIn actions run | A script driving a browser on your machine | Their servers, often behind proxy IPs | The browser you already use. No cloud automation, no proxy IP |
| Where you work | In your terminal | In your terminal | Wherever suits the task: your terminal, the TalkToHumans app, or right inside LinkedIn and Sales Navigator |
| Who presses send | Usually the agent | The agent or a campaign | You. The CLI only drafts |
No LinkedIn tool is completely risk-free. Here is how TalkToHumans keeps the risk down:
- Human pace, enforced. Every action type has its own conservative daily limit and a slightly random gap between actions, so nothing goes out in bursts. Limits belong to the LinkedIn account, so every teammate and every agent shares them. Nobody can raise them, not even your agent.
- Cache-based sync. Your agent reads a synced copy of your LinkedIn data, not LinkedIn itself. Run 100 commands on your inbox and your account sees the same activity as one, with no hopping from profile to profile. The background sync that fills that copy is paced the same way.
- Posts and searches skip your account. Posts, comments, the people engaging with them and people searches come from a data provider, never from your LinkedIn session.
- Runs in your browser. LinkedIn actions go through the browser you already use. No cloud automation, no proxy IPs and no virtual browsers acting on your behalf.
- Work in the terminal, the app or LinkedIn. Everything your agent does also lands in the TalkToHumans app and in panels embedded in LinkedIn's own pages. Check conversations, see what's scheduled and take over by hand wherever you are.
- Drafts only.
tthcan write a reply or stage a full follow-up sequence. No command sends a message: nothing goes out until you hit send.
Founders, consultants, sales teams and agencies use it for these jobs, most common first:
- Fresh leads, ready to contact. Your agent works out who you sell to, finds new leads warmest first (people who engaged with you, posts on your topic, then a LinkedIn search by role, company or location), scores them on your ICP and stages a sequence in your voice for each. You review and launch them from one view.
- Find deals buried in your inbox. Scan hundreds of LinkedIn, Sales Navigator and Recruiter threads for the ones where someone is waiting on you, or showed interest and went quiet, every morning if you like.
- Follow up on silent accepts. Follow-ups for people who accepted but never replied, written for each person. They stop when the lead replies, and wait for you to launch them.
- Log chats in your CRM. Combine it with your CRM's CLI or MCP to log LinkedIn conversations, or use the native HubSpot integration.
- Import a lead list. Bring in a list from an event, a webinar or Sales Navigator, up to 500 profiles at a time, and get a ready-to-work view with first messages drafted.
- Leads already in your network. Describe your ideal customer, and your agent scores your connections. They're already connected to you, so you can message them today.
- Coffees on your next trip. See the people in the city you're visiting on a map, and get coffee invites drafted, to network or sell.
- Leads from post comments. Find posts on a topic or by a competitor, and import the people who commented or reacted, each with a signal saying how.
- Organize your inbox. Set up tags and inboxes that fit your business, then tag hundreds of contacts the way your team already works.
- What works in your outbound. Analyze acceptance rates, reply rates and templates, from connection request to booked meeting, and what to change next.
- Score leads by ICP. Fill an ICP score for every contact, so you can sort them by best fit.
- What works in your content. See which of your posts bring engagement and which bring the people you want to reach, compared with the best and worst posts in your niche.
- Find your quick wins. Ask what you can do today, with little effort, that pays off.
- Script it. Pipe JSON into
jq, run a weekly report from cron, or wire LinkedIn into your own tools.
Your agent also gets work email and mobile phone for the people you talk to, from 20+ providers in a waterfall.
# Install (macOS and Linux)
brew install talktohumans/tap/tth
# Sign in (opens a browser)
tth auth login
# List your contacts
tth contacts list --query "Acme"On Windows or without Homebrew, download tth from the latest release and put it on your PATH.
Scripts and cron jobs sign in with an API key instead: set TTH_API_KEY to a key a workspace admin creates in TalkToHumans Settings.
Install the agent skill once:
tth skill --installOr with any skills installer: npx skills add TalkToHumans/linkedin-cli
# Conversations where you owe a reply
tth conversations list --account-id 42 --filters '[{"field":"conversation.reply_state","operator":"=","value":"waiting_for_my_reply"}]'
# Import profiles and wait for the result
tth contacts import --account-id 42 --file leads.txt --wait
# Find contacts by LinkedIn URL
tth contacts list --account-id 42 --filters '[{"field":"member.linkedin_url","operator":"=","value":["https://www.linkedin.com/in/carina-x/"]}]'
# Enrich and tag a contact
tth contacts enrich --contact-id 123
tth contacts update 123 --tag "Hot lead"
# Draft a message (never sent)
tth drafts write --account-id 42 --contact-id 123 --draft-message "Hi Ada, ..."
# Last week's performance
tth activity summary --from 2026-10-19 --to 2026-10-25Every command supports --help, starting with real examples:
tth contacts --help
tth drafts write --help| Area | Commands |
|---|---|
| Auth | tth auth login, tth auth status, tth auth logout |
| Accounts | tth accounts list, tth credits show |
| Conversations | tth conversations list, tth messages list, tth drafts write, tth files upload |
| Contacts | tth contacts list, import, import-status, update, delete, tth people search, tth signals create, tth fields list, create, tth tags list, update, delete |
| Companies | tth companies list |
| Content | tth content list, search, find, import, import-engagers, track, update |
| Enrichment | tth contacts enrich, find-email, find-phone |
| Views | tth views list, create, update, delete |
| Templates & sequences | tth templates list, create, update, tth sequences update, delete |
| Reporting | tth activity list, tth activity summary |
| Agents | tth skill, tth skills list, tth skills show, tth feedback send |
| Anything else | tth api <METHOD> /v1/... |
tth drafts write saves a draft or stages a sequence for you to review. It never sends. Imports, searches, enrichment and tracking use TalkToHumans credits. tth accounts list --json also gives your agent your own LinkedIn profile, for free, so it knows who you are.
Both reach the same workspace with the same safety rules and the same playbooks. Pick by where your agent runs:
| LinkedIn CLI | LinkedIn MCP | |
|---|---|---|
| Best for | Terminal agents (Claude Code, Codex, Cursor), scripts and cron | Chat apps (Claude, ChatGPT), n8n and any remote MCP client |
| Context cost | Only the commands your agent runs, and --help when it needs it |
Every tool definition, in every session |
| Login | tth auth login in the browser, or an API key |
OAuth when you add the connector |
Many people use both: the MCP in their chat app, the CLI in their terminal. To connect Claude, ChatGPT or another chat app (Claude is one click), follow the MCP setup guide.
A command-line tool for LinkedIn. You, a script or an AI agent run commands like tth conversations list in a terminal and get your LinkedIn data back as a table or JSON, instead of opening LinkedIn in a browser tab.
LinkedIn doesn't offer an official one, and its public API doesn't cover private messages. tth is a third-party LinkedIn CLI by TalkToHumans that works on your own account: conversations, network, contacts, enrichment, drafts and reports.
MCP tool definitions are loaded into your agent's context in every session and every subagent. A CLI costs tokens only when your agent calls it, and its help is read on demand. It also runs in scripts and cron jobs, where no agent is involved.
No, and that's on purpose. Your agent can read, sort and draft, down to a full follow-up sequence. Every message that carries your name still gets your approval.
No. TalkToHumans works through your logged-in browser session. tth signs in to TalkToHumans, not to LinkedIn, and stores its login in ~/.config/talktohumans/, readable by you only.
Your workspace syncs while your browser is open with the TalkToHumans extension on, newest conversations first. Older message history loads when you open a conversation in TalkToHumans. tth messages list warns on stderr when a conversation isn't fully synced.
The CLI, the MCP and the API come with every paid plan, from Starter. Pro adds sequences, post engagers, analytics and CRM integrations.
This repo holds the tth release binaries, the agent skill (skills/talktohumans-linkedin/) and this documentation. The Homebrew formula lives in homebrew-tap. tth talks to the hosted TalkToHumans API at https://api.talktohumans.app/v1, and its source code isn't in this repo.
Found a bug or a missing command? Open an issue, or run tth feedback send.
Built by TalkToHumans.
