LinkedIn outreach
that stops itself.
ReplyMama drives the Chrome profile you are already signed into, spaces its sends like a person would, and holds itself to a daily cap. If LinkedIn raises a flag it stops, and only you can restart it.
- Runs on your Mac
- Your LinkedIn session never leaves it
- Bring your own AI key
- 1LatchA restriction signal outranks every other check.
- 2Daily capToday's point on the warming ramp.
- 3Weekly capLinkedIn's own weekly ceiling for your plan.
- 4WindowLocal working hours, weekdays only.
- 5DelayA jittered gap before the next action is allowed.
| LinkedIn plan | Per day | Per week |
|---|---|---|
| LinkedIn Free | 20 | 100 |
| Premium | 30 | 150 |
| Sales Navigator | 40 | 200 |
Gaps are jittered ×0.8–×1.2 rather than fixed, so the cadence never forms a metronome.
Not a mockup. Every figure here is read from the same constants the engine enforces, and a test fails if they stop matching the code that ships.
The cap is enforced
A hard 25 invitations a day, or less if your LinkedIn plan or a warm-up period sets it lower. Nothing in the app raises it.
You approve every message
The AI drafts from the lead's profile and puts the note in a queue. It is sent after you approve it, not before.
Nothing leaves your Mac
The app drives your own Chrome profile locally. We never see your cookies, your password or your session token, so there is nothing of yours to lose.
Find, draft, send, stop.
- 01
Find
Paste a LinkedIn search URL and the app pages through the results and saves the leads. It only reads. If you already have a CSV, import that instead.
- 02
Draft
Connect your own OpenAI-compatible key. The app writes a note for each lead from what is on their profile, then puts it in the approval queue.
- 03
Send
Approved notes go out through your own browser, 90 seconds to 5 minutes apart, during working hours. When someone accepts, their follow-up moves up the queue.
- 04
Stop
A reply pauses that lead's sequence. A restriction signal stops everything. Neither one clears itself on a timer.
Hard limits, enforced in the engine.
- The daily cap holds at 25
- Your LinkedIn plan or a warm-up period can hold it lower, never higher. Once the day's count is reached the queue stops and the app tells you when it resumes.
- A stop stays stopped
- When LinkedIn raises a flag the app writes the stop to disk and stops sending. Quitting, reinstalling or rebooting will not clear it. You clear it from inside the app, once the cause is resolved.
- Nothing sends without your approval
- The AI writes the first draft and puts it in a queue. It goes out after you approve it, never on the schedule alone.
- Your LinkedIn session stays on your Mac
- The app drives the Chrome profile you are already signed into. We never receive your cookies, your password or your session token, so there is nothing on our side to lose.
Before every send the app checks for five things: an invitation limit, a weekly ceiling, an unusual-activity warning, a security checkpoint, or a captcha. Any one of them stops sending until you have cleared it.
Everything it does.
Lead capture
- Search-to-leads scraping
- CSV lead import
- Per-campaign lead lists
Outreach
- BYOK AI drafting with approval
- Accept detection and follow-ups
- Stale-invite withdrawal, dry-run first
Visibility
- Per-campaign funnel analytics
- Unified inbox with reply-stop
- Public API and webhooks
Operations
- Durable restriction stop
- Team invites and roles
- White-label branding
Straight answers.
Restrictions come from accounts that behave like bots: constant volume, instant actions, activity at 3am, and no reaction when LinkedIn pushes back. ReplyMama is built to do the opposite, and the behaviour is enforced in the engine rather than left to your judgement. A hard daily ceiling, a slow ramp on a new account, human-length gaps between actions, sending only inside your working hours, and an immediate stop when LinkedIn signals anything unusual. Those are the protections that keep accounts safe in practice.
No. Your LinkedIn session never reaches us. The app drives your own installed Chrome on an isolated profile, so the cookies, tokens and passwords stay on your Mac and are never collected or transmitted. What we hold is your login to this site, your campaigns and your billing. That separation is the whole architecture: there is no relay to leak.
Everything stops, immediately and durably. The stop is written to disk inside your profile, so a crash, a restart or a reinstall will not clear it. Nothing resumes on a timer. You clear it deliberately, from the app, once you have dealt with the cause. That is what makes the stop trustworthy rather than decorative.
ReplyMama runs on your Mac, so it sends while your Mac is awake and working. That is the design: every send happens on a machine you control, through a browser you are already signed into. There is no cloud worker sending in the background on your behalf, which is exactly why there is so little of yours to expose. If you want outreach running round the clock from a server, a hosted tool will suit you better.
Templates fill blanks. ReplyMama reads what is actually on the profile, such as recent posts, a role change or company context, and writes a note from scratch, so it reads like something a person wrote. Then it waits for your approval, so nothing goes out you have not read. You bring your own model key, which keeps the cost on your account and keeps prospect text out of a model we pay for.
Sequences pause and stay paused, and nothing auto-resumes. Your leads, campaigns and history stay in your account so you can pick up where you left off, and you can delete the workspace whenever you want. Deletion is a real cascade, not a flag on a row.
Same safety on the cheapest plan.
Every plan ships the same caps, the same approval queue and the same stop button. Plans differ on seats, LinkedIn accounts and white-label.
Flat monthly pricing. Cancel any time.