Sales outreach · checked 2026-10-19Can I vibecode Omentir?
NOT REALLY· don't bother
The campaign engine is easy to describe: a lead list, message templates, a daily cap and a reply inbox. What no build can supply is LinkedIn itself. Omentir sends invites and messages from your own account through a provider, and its open-source repo needs Firebase, an AI key and a Unipile account, whose entry price is about as much as the $49 subscription. That provider works by reverse engineering LinkedIn, and LinkedIn's user agreement restricts bots and scraping, so an account can be restricted whoever wrote the code. What you can build and keep is the desk around it: leads, scoring, drafted notes you send yourself and reply sorting.
$49/moprice
$588it costs you a year
one sittingbuild time
—people replaced it
What you need on hand.
DifficultyBeginner · No setup beyond your agent.
TimeOne sitting for the desk; the LinkedIn automation itself can't be built safely
- Claude Code, Codex or CursorAny of them can run the prompt.
- Node.js 22better-sqlite3 compiles a native module, so build tools must be installed.
- An LLM API keyAny OpenAI-compatible endpoint works, including a local model.
- A list of leadsCollect them yourself; the tool never touches LinkedIn.
Before you buildOmentir has no free hosted plan. Self-hosting its MIT-licensed code needs Firebase, an AI key and a Unipile account (from about €49 a month), so it isn't a cheaper route.
- 1. Install one agent: Claude Code, Codex, Cursor.
- 2. Make an empty folder and open it in the agent.
- 3. Paste the prompt below, press enter, and answer its questions as it builds.
The prompt
Build me a LinkedIn prospecting desk that I operate by hand, covering the part of Omentir I can build without automating LinkedIn. Requirements: - Node.js 22 + TypeScript: Fastify with htmx pages on http://127.0.0.1:4910; storage in better-sqlite3 at ./data/prospects.db. - Leads: import a CSV (papaparse) with name, title, company, company size, location, profile URL and a free-text "recent activity" column; dedupe by profile URL; add tags and notes. - ICP: a settings page where I describe my ideal customer (titles, industries, company sizes, exclusions). Score each lead 0-100 with an OpenAI-compatible model (OPENAI_BASE_URL, OPENAI_API_KEY and MODEL in .env), as JSON with a one-line reason checked by zod, and cache the result per lead. - Drafts: for a chosen lead, write a connection note within a limit set in .env (default 200 characters, because LinkedIn's limit varies by account type) and one follow-up, using my offer text and the lead's row only. Show the character count and never invent facts that aren't in the row. - Queue: a Today page listing up to N leads to contact (N in .env, default 20), each with an "Open profile" link, a copy button for the note, and buttons sent, skip and replied. I send everything myself inside LinkedIn. - Replies: paste a reply into a lead's page; the model labels it interested, not now, wrong person or stop and drafts a response I can edit; "stop" adds the lead to a do-not-contact list. - Pipeline: stages new, invited, connected, replied, meeting and lost, a weekly count per stage, and CSV export of everything. - No accounts, no telemetry, local-first: only the model call leaves the machine, and the README says what lead text is sent. - Out of scope: any automated sending, scraping or browser automation on LinkedIn, lead search, cold email. - README: the .env keys, LinkedIn's rules against automated activity and why this tool never touches the site, and GDPR basics for storing other people's data.
Opens your agent with the prompt filled in, you press enter. Or copy it and paste it into an empty folder.
Then keep going
Next prompts, same folder.
The first prompt builds the core. Run these one at a time in the same folder to grow it into your own Omentir.
Add a weekly review page to the prospecting desk in this folder. Requirements: - Per ISO week, counts of leads added, notes marked sent, replies, meetings and losses, with the conversion between each stage. - Break replies down by the label the model gave them, and by the tag or ICP score band of the lead, so I can see whether high scores really reply more. - A table of leads stuck in a stage for more than 10 days. - Export the review as Markdown. Update the README.
Add company research notes to the prospecting desk in this folder. Requirements: - On a company page, a text box where I paste an About page or recent post; the model returns 3 short hooks that quote only that text, shown for my approval before they're saved. - Drafts for any lead at that company can use the approved hooks, and the draft shows which hook it used. - Never fetch pages automatically; every piece of text is pasted by me. - Update the README.
Add note templates with results to the prospecting desk in this folder. Requirements: - Save named templates with variables like {{firstName}} and {{company}}, and let me pick one as the base for a draft; keep a version history. - Record which template each sent note used, and show sent, accepted and replied counts per template once a template has been used at least 20 times. - Say "too few to compare" below that number; never show a percentage without the count. - Update the README.
What you lose
- Invites and follow-ups sent for you on a schedule from your LinkedIn account
- Lead search against LinkedIn with fit scoring
- Targeting of people who comment on competitors' posts
- A live inbox of LinkedIn replies and the MCP and REST API for Claude, ChatGPT and Cursor
- The three-bookings-a-week guarantee
Why people still pay
Founders pay $49 to skip hiring an SDR and stitching a lead database, a sequencer and an inbox together, and the refund guarantee lowers the cost of trying. Omentir says it paces sending at a human rate with limits you set, but its own FAQ and help pages say LinkedIn's rules still apply and automation carries account risk.
Omentir or your own copy?
Over
Your usage
3h ×$/h
Omentir, 1 yr$588
Your own copy, 1 yr$65running + build session
You keep$523pays for itself in the first month
Estimate · Runs on your machine. LLM tokens are the only running cost, nothing with a local model. This covers the desk, not automated sending. The build is token usage on your agent plan.
Take your data, then cancel.
1 · Take your data out of Omentir
- 1Omentir's site and pricing page describe no export of leads, campaigns or inbox.
- 2Pro includes API access, and the README links an OpenAPI description of the agent API at omentir.com/api/agent/v1/openapi.json: use it to read leads and inbox threads yourself.
- 3Its terms say you are responsible for exporting or preserving information before access changes, and that a refund can end access to stored outputs.
- 4Your LinkedIn connections and conversations stay in LinkedIn itself.
2 · Cancel Omentir
- 1The terms say you may cancel at any time to stop future renewals, but they don't describe the steps.
- 2We found no help article on cancelling: use the billing page inside the app, or email hi@omentir.com, the address the guarantee policy gives for support.
- 3Disconnect the LinkedIn account in the app before you stop, so no queued outreach keeps running.
- Standard refund requests must come within 7 days of purchase and are paid minus payment fees; the booking guarantee has its own 14-day request window.
- Cancelling or disconnecting mid-week can cost you eligibility for that week's guarantee.
- Omentir's own help pages warn that LinkedIn can restrict accounts that use automation: check your account's health before and after.
The escape hatch
Don't feel like building it? These are already free.
Omentir (self-hosted)Open sourceThe MIT-licensed code behind the paid product, run with Docker; it needs your own Firebase, Unipile and Gemini or Vertex AI accounts, so it isn't dependency-free or free to run.★ 47 · active Oct 2026 · docker to self-host
WebVisit ↗LinkiFreeSelf-hosted AI SDR for LinkedIn sequences, cold email and lead enrichment; source-available under its own Sustainable Use License rather than an open-source one.★ 168 · active Oct 2026 · docker to self-host
WebVisit ↗
WebVisit ↗LinkiFreeSelf-hosted AI SDR for LinkedIn sequences, cold email and lead enrichment; source-available under its own Sustainable Use License rather than an open-source one.★ 168 · active Oct 2026 · docker to self-host
WebVisit ↗
The numbers
Omentir pricing
| Plan | Monthly | Annual, per month | What you get |
|---|---|---|---|
| Pro | $49 | — | 1 user, 1 LinkedIn account, unlimited AI agents, leads and campaigns, AI-automated campaigns, API access, email support. Carries a booking guarantee: 3 bookings a week or a refund. |
| Enterprise | custom | — | Unlimited users and LinkedIn accounts, SSO, dedicated onboarding, priority support, managed campaigns. |
- Billing
- Priced at $49 a month and paid through a payment provider; the pricing page doesn't mention a trial or yearly billing, and the terms don't state the billing frequency. The booking guarantee has a two-week warm-up with no guarantee, needs an active agent for the whole week and pays back the current subscription charge minus payment fees (usually 3.5% to 5%), on request within 14 days. Otherwise refund requests must come within 7 days of purchase.
- Hidden costs
- Sending limits come from LinkedIn and from your own Settings, and the FAQ says you still own compliance with LinkedIn's rules. The terms say Omentir doesn't guarantee account safety and isn't affiliated with LinkedIn. Extra LinkedIn accounts and users need Enterprise.
- Source
- omentir.com
Also on the list
Related apps
No free hosted plan. The code is MIT-licensed and can be self-hosted, but it needs Firebase, Unipile and an AI provider, which cost money.
Not really. Founders pay $49 to skip hiring an SDR and stitching a lead database, a sequencer and an inbox together, and the refund guarantee lowers the cost of trying. Omentir says it paces sending at a human rate with limits you set, but its own FAQ and help pages say LinkedIn's rules still apply and automation carries account risk. The prompt on this page builds the closest honest substitute.
About $49 a month on the Pro plan, $588 a year (checked 2026-10-19).
Invites and follow-ups sent for you on a schedule from your LinkedIn account; Lead search against LinkedIn with fit scoring; Targeting of people who comment on competitors' posts; A live inbox of LinkedIn replies and the MCP and REST API for Claude, ChatGPT and Cursor; The three-bookings-a-week guarantee. If any of those matter to you, keep paying.
Yes: Omentir (self-hosted) (open source), Linki (free).