The loop, step by step

Your week goes in. Posts that sound like you come out. You approve every one.

Liftli is a loop, not a text box. It gathers your real work, drafts in your extracted voice, waits for your one-tap yes, and learns from what the post actually did. Here's each step, in depth.

Updated July 2026

TL;DR

Connect once inside Claude (about 2 minutes; Claude Code too, with ChatGPT, Codex, Cursor, Hermes and OpenClaw in beta). Liftli gathers raw material from your voice notes, calls, commits and the news in your lane; runs each idea through a plan → draft → critique → revise loop in your extracted voice; then stops at a one-tap gate. You approve, tweak, or skip. It logs what happened and recalibrates next week, so it needs you less each time. Nothing ever touches your accounts through extensions, scraping, or bots.

01Connect

Connect once.

Liftli lives inside the AI you already use — Claude today (a free Claude account is enough) and Claude Code, with ChatGPT, Codex, Cursor, Hermes and OpenClaw in beta. There's no dashboard to learn and nothing to install on LinkedIn.

Setup takes about two minutes: the Start free flow walks you through connecting Liftli to Claude and pulling in a sample of your writing so it can extract your voice. From then on, you talk to your AI the way you already do — Liftli is just there, holding your strategy and memory.

The free tier is a one-time lifetime allowance, no card: 3 posts (edit each one as much as you like), 10 comments, 3 idea extractions, and voice extraction. Enough to see whether the drafts actually sound like you before you pay anything. Plans are on the pricing section.

02Capture

It gathers your actual week.

Most tools start you at a blank topic box. Liftli starts at your actual week. It pulls raw material from:

  • Voice notes — talk into your phone via Telegram; a 90-second ramble on the drive home is a post seed.
  • Call transcripts — the sharp thing you said in Tuesday's client call, captured instead of lost.
  • GitHub activity — what you actually shipped, not what you vaguely remember shipping.
  • Chats — the explanation you typed to a colleague that deserved a bigger audience.
  • The news in your lane — so you can react while a story is still a story.

This is why Liftli drafts don't read like everyone else's: the input is your life, not a shared inspiration library. Detectably-AI posts underperform human writing in most professional niches (Originality.AI's 2025 study of 3,368 posts) — the fix isn't better templates, it's better raw material.

A da Vinci ornithopter flying machine folded from kraft and cream paper, with paper gears and struts
03Craft

It drafts in your voice.

Every draft is written in your extracted voice — patterns learned from your own writing, kept persistently per person — and matched to your strategy: your positioning, your audience, your calendar, what's worked before.

And no draft goes straight from idea to your screen. Each one runs a full plan → draft → critique → revise loop first: Liftli plans the angle, writes it, critiques its own work against your voice and strategy, and revises — before you ever see it. What lands in front of you is the version that survived that process, not a first attempt.

The same engine drafts comments for posts worth engaging with, so your presence isn't just publishing — it's participating.

04The gate

You say yes.

Liftli's pipeline runs on autopilot up to a one-tap gate: approve, tweak, or skip. Early on you will read everything — that is how it learns your judgement, not just your voice. The better it knows you, the less you change, and the more you hand it outright.

Why insist on the gate?

  • Feeds punish unsupervised AI. Detectably-AI content underperforms human writing in most professional niches. A human decision on every post is what keeps your content on the right side of that line.
  • Your name is the asset. One post you'd never have said, published under your name, costs more than a hundred good ones earn. The gate means that can't happen.

The gate is also cheap: reading a draft and tapping yes takes seconds. You keep the judgment; Liftli keeps the labor.

05Memory

It learns.

After you post, Liftli closes the loop. It logs outcomes, scores engagement on what you published, and surfaces warm leads — the people whose engagement patterns suggest they're worth a conversation, not just a like (lead analysis up to 100 leads on Pro).

Then it recalibrates: next week's plan reflects what this week's posts actually did. Which angles landed, which formats fell flat, who showed up. Your skips and tweaks feed the same memory — the drafts drift toward you, not away. That persistent strategy-plus-memory is the difference between a writing tool and a strategist.

One founder data point on where this can lead: a Liftli-drafted LinkedIn comment became a cited source in AI search results — discovered by accident by Liftli's founder, Oded Tsamir. LinkedIn is the #1 cited source for professional questions in AI search, per Profound's 2026 citation reports. More on that on LinkedIn and AI citations.

06Breakthrough

Publishing, the safe way.

Today, publishing is deliberately simple: Liftli hands you the approved text and you post it yourself — copy, paste, publish, on LinkedIn, X, or Substack. Your accounts stay untouched.

Native publishing and scheduling work one way only: through the platforms' official APIs, connected only when you explicitly link an account. Never through browser extensions, scraping, or bots acting on your account — the category LinkedIn's User Agreement prohibits and has enforced against (extension tools restricted, and Kleo, with 70K users, shut down mid-2025). If that risk is why you're shopping, the Taplio alternative page goes deeper.

If you already run a scheduler you like, keep it — Liftli sits alongside one happily. You don't need one, though: Liftli publishes and schedules to LinkedIn and X itself (see Typefully vs. Liftli).

07The honest part

When this loop is not for you.

Honesty over conversion. If you don't use an AI assistant day to day, you'll be adopting two things at once. And if what you really want is a standalone scheduling dashboard with queues and team seats, a dedicated execution app is the better buy — check current details on each vendor's site.

Liftli is for people whose week is already full of good material and whose bottleneck is turning it into a consistent, in-their-own-voice presence — founders, consultants, builders. That's the loop above, run every week.

08FAQ

Questions, answered straight.

What do I need to use Liftli?
An account with one of the AIs Liftli runs in — Claude, Claude Code, Grok or Grok Bot today, with ChatGPT, Codex, Cursor, Hermes and OpenClaw in beta. On Claude the free tier works. That's it: no browser extension, no separate app to learn, nothing installed on your LinkedIn account. The free tier needs no credit card; the Start free flow at app.liftli.ai walks you through the two-minute connection.
How much of my time does Liftli take each week?
The raw material is work you already do — voice notes you talk into your phone, calls you were on anyway, commits you already pushed. Liftli does the gathering, planning, drafting, critiquing and revising before you see anything. Your job is the decision: read the draft, then approve, tweak, or skip. Most of the loop runs without you in it.
What happens if I hate a draft?
Skip it or tweak it — nothing publishes without your one-tap yes, so a bad draft costs you a tap, not a bad post under your name. Every skip and edit feeds back into Liftli's persistent memory, so the next round of drafts moves closer to what you'd actually say.
Which platforms does Liftli cover?
LinkedIn, X, and Substack. Approved posts publish or schedule straight from chat through the platforms' official APIs — an account is connected only when you explicitly link it, never through extensions, scraping, or bots.

The loop starts with a two-minute connection.

Connect Liftli inside Claude, talk into your phone this week, and approve the first draft it hands back.

A clay-colored paper airplane on a launch stand, angled for takeoff on a drafting table marked with red chalk