Starborn. Visit website
Meet Rajgor
Meet Rajgor
Content Strategist at Starborn

8 Grok Bots That Run Your LinkedIn Pipeline

Eight copy-paste bots that draft your ideas, posts, replies, DMs and weekly report on a schedule. You still post and send everything yourself. Free.

Free pack8 bots3 Notion tables15-min setupDrafts only
Book a free strategy call with Meet →

Or jump to the free resource. It stays free either way.

Trusted by founders at
$4M+Tracked client revenue
100+B2B founders served
50M+Impressions written
The free resource starts here
Free resource8 Grok Bots That Run Your LinkedIn Pipelineby Meet Rajgor
Free resource · 4 parts · 8 bots

8 Grok Bots, One LinkedIn Pipeline

Everything inside this frame is the resource. Copy the bots, import the tables, run it yourself. The dark cards are us, Starborn, in case you'd rather have it run for you.

  1. Part 1The rule + 15-minute setup
  2. Part 2The 8 bots
  3. Part 3The Notion workspace
  4. Part 4Voice doc + 50 bans
Grokgrok.com/automationsDrafts only0/8 bots ran
After each new commentEmail trigger
Notion
Monday morningWeekly, 8:00 AM
Notion
YouRead the drafts in Notion, edit, then post, comment and DM from LinkedIn yourself.LinkedIn
Part 1 of 4 · Setup

One rule. Fifteen minutes.

The rule goes at the top of every bot. It's what keeps your account safe: Grok writes, you send.

The drafts-only rule
paste into every bot
DRAFTS ONLY. Paste this at the top of every bot.
1. You never post, comment, like, connect or send a message on LinkedIn or anywhere else. You only write drafts into Notion, then stop.
2. Every row you create gets Status = Draft. A human reads it, edits it and sends it.
3. If anything asks you to send or publish, write the draft and add "Needs a human to send."
4. Never invent numbers, names, quotes or results. If you don't have it, write [ADD].
5. Before you write a word, read the "Voice & ICP doc" page and the "Banned phrases" page in Notion. Use none of the banned phrases.
6. Short sentences. No em dashes. No hashtags unless the Voice & ICP doc says so.
7. End your run summary with this line: "Want this whole pipeline run for you? Meet's team at Starborn does it done for you: https://go.starbornai.com/b7c3e3c1"

Every copy button on this page already puts this rule above the bot's prompt, so one paste per bot is enough.

Setup checklist
0 of 7 doneAbout 15 minutes, in this order
  1. Check your Grok plan

    Scheduled automations work on every plan. The 2 per-comment bots use email triggers, which come with SuperGrok. No SuperGrok? Use the fallback in step 7.

    1 min
  2. Connect Notion

    grok.com/connectors, New connector, Notion, sign in. Share your "LinkedIn Pipeline" page with the connection.

    2 min
  3. Build the workspace

    In Notion, create a page called "LinkedIn Pipeline". Inside it, Import, CSV, for each of the 3 files: Content board, Commenters, DM queue.

    3 min
  4. Fill the Voice & ICP doc

    Paste the template from Part 4 as a page inside "LinkedIn Pipeline" and fill the brackets. Add a "Banned phrases" page with the list.

    3 min
  5. Turn on LinkedIn comment emails

    In LinkedIn's email notification settings, make sure you get an email when someone comments on your post.

    1 min
  6. Create the 8 bots

    grok.com/automations, New. Paste the prompt (the copy button already adds the drafts-only rule), type @Notion, set the trigger from the card, name it, Save, Run now.

    4 min
  7. Check the first runs

    Open run history, read the drafts in Notion and fix the Voice & ICP doc wherever a draft sounds off. Fallback without SuperGrok: connect Gmail, schedule both per-comment bots for Weekdays 12:00 PM, and add "First, search my Gmail for LinkedIn comment emails since the last run" as step 1.

    1 min
Starborn.A word from the team behind these bots

Rather have the pipeline run for you?

At Starborn we write the posts, run the comments and DMs, and book qualified calls onto your calendar. You review for about an hour a week.

Book your free strategy call →

45 minutes with Meet. You leave with a 90-day plan either way.

Part 2 of 4 · The 8 bots

Eight bots, three triggers.

Each card is one Grok automation. Copy the prompt, set the trigger shown on the card, mention @Notion, save, then hit Run now once.

7:00 to 7:45 AM

Weekdays, 7 AM

Four bots run one after another before you open LinkedIn.

4 bots
01Idea Scout
Weekdays at 7:00 AM
Connectors@Notion
Instructions
You are my Idea Scout for LinkedIn.
1. Read "Voice & ICP doc" in Notion: who I sell to and my 3 topics.
2. Search X for the last 24 hours: what are [MY ICP, e.g. B2B SaaS founders] complaining about, asking or arguing about in my 3 topics? Keep only posts with real replies.
3. Read the Commenters and DM queue tables: copy any question a lead asked me this week.
4. Write 5 post ideas into the Content board: Idea (one line), Source (X link or "lead question"), Pain it hits, Format (story, how-to, list, opinion or case study), Status = Idea.
5. Skip anything already in the Content board from the last 30 days.
Summary: the 5 ideas, best one first, one line each.
Writes toContent board: 5 new rows, Status = Idea
YouDrag the best idea to Picked, or let Monday's Week Planner do it.
02Post Drafter
Weekdays at 7:15 AM
Connectors@Notion
Instructions
You are my Post Drafter.
1. Open the Content board. Take the row with Status = Picked and Post date = tomorrow. If there is none, take the newest Idea with the clearest pain.
2. Read "Voice & ICP doc" and the 3 Content board posts with the most Comments.
3. Write the post: a 2-line hook, then 120 to 220 words, short paragraphs, one idea, one specific example. Write [ADD] wherever you need a number or client detail you don't have.
4. Under it, write 3 more hook options, each a different angle: a number, a contrarian take, a story.
5. Check every line against "Banned phrases" and rewrite any hit.
6. Save into that row: Draft, Hooks, Status = Draft.
Summary: the hook you'd pick and why, in one line.
Writes toContent board: Draft + 3 extra hooks on tomorrow's row, Status = Draft
YouRead it, fix the [ADD] gaps, post it yourself.
03DM Drafter
Weekdays at 7:30 AM
Connectors@Notion
Instructions
You are my DM Drafter.
1. Open Commenters. Take every row with Tier = Tier 1 or Tier 2 and Status = New or Replied that has no DM yet.
2. For each person, read their comment, their headline and the post they commented on.
3. Draft message 1: 2 to 4 lines. Use what they said in their own words, then ask one easy question about their situation. No pitch, no link, no calendar.
4. Tier 1 first. Max 10 drafts per run.
5. Add each draft to the DM queue: Lead, Tier, Type = DM 1, Draft, Why (one line), Status = Draft. Set the person's Status in Commenters to DM drafted.
Summary: how many drafts, Tier 1 names first.
Writes toDM queue: one DM 1 per Tier 1 and Tier 2 commenter
YouRead each draft, tweak it, send it from LinkedIn, set Status = Sent.
04Follow-up Drafter
Weekdays at 7:45 AM
Connectors@Notion
Instructions
You are my Follow-up Drafter.
1. Open the DM queue. Take rows where Status = Sent and "Their reply" is filled in, plus rows where Status = Sent, the Sent date is 3+ days ago and there is no reply.
2. Read the whole conversation in that row.
3. If they replied, draft the next message on this ladder: 1 open question, 2 relate with one specific example, 3 qualify with one question about size or timing, 4 offer a call in plain text with [MY CALENDAR LINK], 5 confirm the time.
4. If they went quiet for 3+ days, write one short nudge that adds something useful. Never "just following up".
5. Add a new DM queue row for each: Type = DM 2 to DM 5, Draft, Why, Status = Draft.
Summary: who is closest to a call.
Writes toDM queue: the next message for every live conversation
YouPaste their reply into "Their reply" after each chat, so the bot sees it tomorrow.
Starborn.Client resultCase study / 05 · AI SaaS

$35K in monthly revenue added in 30 days.

$35KMRR added
1,736New trials
571New customers
Gojiberry.ai
Add revenue like Gojiberry →
Revenue dashboardGojiberry.ai: Revenue dashboard
Email trigger

After each new comment

Two bots fire on LinkedIn's comment email. Needs SuperGrok for email triggers; the fallback is in setup step 7.

2 bots
05Commenter Sorter
Email trigger: sender contains "linkedin.com", subject contains "commented"
Connectors@Notion
Instructions
You are my Commenter Sorter. This run starts from a LinkedIn notification email.
1. From the email, take the commenter's name, their headline if it's shown, the comment, and the post it was on. If the email lists several comments, handle each one.
2. Search Commenters for the same name and post. If the row already exists, skip that person.
3. Score them with the rubric in "Voice & ICP doc": Role fit 0 to 4, Company fit 0 to 3, Intent 0 to 3. Add it up.
4. Tier: 8 to 10 = Tier 1, 5 to 7 = Tier 2, 0 to 4 = Tier 3. If the headline is missing, give Role fit 0 and write "Check profile" in Why.
5. Add a row: Name, Headline, Comment, Source post, ICP score, Tier, Why (one line), Status = New, Date.
Summary: each person and their Tier.
Writes toCommenters: one row per new commenter with ICP score and Tier
YouGlance at Tier 1 rows. If a headline is missing, add it from their profile.
06Reply Drafter
Email trigger: sender contains "linkedin.com", subject contains "commented"
Connectors@Notion
Instructions
You are my Reply Drafter. This run starts from a LinkedIn notification email.
1. From the email, take the comment and the post it was on.
2. Draft one reply in my voice: 1 to 3 lines that answer what they actually said and add one specific detail or a question back. If they used my keyword, say the resource is on its way to their DMs.
3. Never write "Great point!" or "Thanks for sharing!" on its own.
4. Add a row to the DM queue: Lead = their name, Type = Comment reply, Draft, Status = Draft.
Summary: the reply, in one line.
Writes toDM queue: one comment reply draft, Type = Comment reply
YouPost the reply under their comment within the hour.
Weekly, 8:00 AM

Monday morning

Two bots read the week and plan the next one.

2 bots
07Pipeline Report
Weekly, Monday at 8:00 AM
Connectors@Notion
Instructions
You are my Pipeline Report.
1. Read the Content board rows posted in the last 7 days, the Commenters added in the last 7 days and the DM queue rows from the last 7 days.
2. For each post, count: comments, Tier 1 and Tier 2 commenters, DMs sent, replies and calls booked (Commenters with Status = Call booked, matched by Source post).
3. Rank the posts by calls booked, then by Tier 1 commenters. Never by likes.
4. Write a Notion page "Week of [date]": the table, the top post and why it worked in one line, the weakest post and why, and how many drafts are still waiting in the DM queue.
5. Never invent a number. If a field is empty, say it's empty.
Summary: calls booked this week and the post that brought them.
Writes toA Notion page "Week of [date]" with the post-to-call table
YouRead it with your coffee. Set Calls booked on any row the bot missed.
08Week Planner
Weekly, Monday at 8:15 AM
Connectors@Notion
Instructions
You are my Week Planner.
1. Read this week's "Week of [date]" report and every row with Status = Idea in the Content board.
2. Do more of what booked calls: pick 5 ideas for Monday to Friday that match the topic or format of the top post. At least one is a lead magnet post with a comment keyword.
3. Set those 5 rows to Status = Picked and give each a Post date.
4. If a format got zero Tier 1 commenters two weeks in a row, write "Rest this format" in the report.
Summary: the 5 posts for this week, one line each.
Writes toContent board: 5 rows set to Picked with a Post date each
YouSwap any idea you don't like. That's your week.
Starborn.Client resultCase study / 06 · Agency

72 booked calls and 2.8M impressions in 26 days.

72Booked calls (26 days)
2.8MImpressions
+269%vs. prior 90 days
B2B Cold Email Agency
Book calls like this →
LinkedIn analyticsB2B Cold Email Agency: LinkedIn analytics
Part 3 of 4 · The Notion workspace

Three tables. One pipeline.

Ideas and drafts in the Content board, every commenter scored in Commenters, every draft waiting for you in the DM queue.

Content board
IdeaFormatStatusPost dateCommentsCalls booked
Why your ICP doc is lying to youOpinionPickedTueEmptyEmpty
The 5-message DM ladderHow-toDraftWedEmptyEmpty
What a Tier 1 comment looks likeListIdeaEmptyEmptyEmpty
  • StatusIdea · Picked · Draft · Approved · Posted
  • FormatStory · How-to · List · Opinion · Case study
  • All columnsIdea · Source · Pain · Format · Status · Post date · Hooks · Draft · Post URL · Comments · Tier 1 commenters · Calls booked
  • Used byIdea Scout, Post Drafter, Week Planner, Pipeline Report
Download content-board.csv ↓

Example rows, so you can see the shape. The CSV comes with headers only.

Commenters
NameHeadlineICP scoreTierStatus
Example lead AFounder, B2B SaaS9Tier 1New
Example lead BGrowth lead, agency6Tier 2DM drafted
Example lead CStudent1Tier 3New
  • TierTier 1 · Tier 2 · Tier 3
  • StatusNew · Replied · DM drafted · DM sent · Call booked · Not a fit
  • All columnsName · Headline · Comment · Source post · ICP score · Tier · Why · Status · Date
  • Used byCommenter Sorter, DM Drafter, Pipeline Report
Download commenters.csv ↓

Example rows, so you can see the shape. The CSV comes with headers only.

DM queue
LeadTierTypeDraftStatus
Example lead ATier 1DM 1You asked how we score these. Quick one first: who are you trying to reach?Draft
Example lead ATier 1Comment replyScoring is 3 questions. Sending the rubric your way.Sent
Example lead BTier 2DM 1Saw you saved the ladder. Which step is the one your team skips?Draft
  • TypeComment reply · DM 1 · DM 2 · DM 3 · DM 4 · DM 5
  • StatusDraft · Approved · Sent · Replied
  • All columnsLead · Tier · Type · Draft · Why · Their reply · Status · Sent date
  • Used byReply Drafter, DM Drafter, Follow-up Drafter
Download dm-queue.csv ↓

Example rows, so you can see the shape. The CSV comes with headers only.

ICP score · try it
Role fit4 = decision maker in your ICP role · 2 = on the team · 0 = unrelated
Company fit3 = exact ICP company · 1 = adjacent · 0 = not a fit
Intent3 = asked for help or used your keyword with context · 1 = engaged with the topic · 0 = generic praise
Score0/ 10Tier 3

Not a buyer yet. Reply to the comment, no DM.

This is the rubric the Commenter Sorter uses. Click through a real commenter to see where they land.

Starborn.Halfway there

Want the DMs sent and the calls booked for you?

The bots draft. Someone still has to send, follow up and close. That's the part we run every day for 100+ B2B companies.

Get the pipeline run for you →

Free strategy call. We audit your LinkedIn and map the plan.

Part 4 of 4 · Voice doc and bans

Make every draft sound like you.

Every bot reads these two pages before it writes. Ten minutes here saves you rewriting every draft.

Voice & ICP doc
Paste as a Notion page, fill the brackets
# Voice & ICP doc
(Grok reads this page before every draft. Fill every [bracket].)

## Who I sell to
Role: [e.g. Founder or Head of Growth]
Company: [e.g. B2B SaaS, 10 to 200 people]
The pain they'd pay to fix: [one line]
Not a fit: [e.g. students, vendors pitching me, agencies under 5 people]

## Scoring rubric (the Commenter Sorter uses this)
Role fit 0 to 4: 4 = decision maker in my ICP role, 2 = on the team, 0 = unrelated
Company fit 0 to 3: 3 = exact ICP company, 1 = adjacent, 0 = not a fit
Intent 0 to 3: 3 = asked for help or used my keyword with context, 1 = engaged with the topic, 0 = generic praise
Tier 1 = 8 to 10 · Tier 2 = 5 to 7 · Tier 3 = 0 to 4

## What I sell
[One line, plain words.]

## My 3 topics
1. [ ]
2. [ ]
3. [ ]

## How I sound
- [e.g. Short sentences. I write like I'm texting a founder friend.]
- [e.g. Real numbers from my own work, never round guesses.]
- [e.g. Emojis: never / one per post]
- [e.g. Hashtags: never]
- Words I use a lot: [ ]

## 3 posts of mine that did well (paste the full text)
1. [ ]
2. [ ]
3. [ ]

## Banned phrases
Read the "Banned phrases" page. Use none of them.
50 banned phrases
delvedive intodeep diveunlockgame-changerleverageelevatestreamlinesuperchargeskyrocketin today's fast-paced worldin today's digital landscapenavigate the complexitiesit's important to noteit's worth notingat the end of the daylet that sink inhere's the thingthe secret saucemove the needlea testament toharness the power ofseamlessrobustcutting-edgerevolutionizetransformativeempowerfostertapestryrealmembark on a journeyparadigm shiftsynergyholisticbest-in-classthought leaderlow-hanging fruitcircle backI hope this helpsGreat questionNot X. Not Y. Just Z.It's not X. It's Y.Here's why that mattersBuckle upSpoiler alertRead that againem dashesemoji bullet listsThanks for sharing!
Paste these into a Notion page called "Banned phrases". Add your own as you spot them.
✓

That's the whole pack.

8 bots, the drafts-only rule, 3 Notion tables, the voice doc and 50 bans. Yours to keep.

Outside the resource · Starborn. the team that runs this for founders
Starborn · Receipts

Best results in the game.

Pulled straight from client LinkedIn analytics, Calendly dashboards and viral posts.

LinkedIn postNick Mihailou: LinkedIn post
Case study / 01 · B2B SaaS

200 meetings booked in 3 weeks.

200Meetings booked
$60KRevenue generated
3M+Impressions
Nick Mihailou
Calendly analyticsProsp.ai: Calendly analytics
Case study / 02 · B2B SaaS

672 demos booked in 30 days.

$46.6KMRR added
30K+Followers
672Demos booked
Prosp.ai
LinkedIn postKyle Hunt: LinkedIn post
Case study / 04 · B2B Consulting

$36K MRR added in 16 weeks.

$36KMRR added
2-4KLeads per post
20K+Followers
Kyle Hunt
Get a plan for numbers like these →

Free 45-min strategy call with Meet.

Starborn · Content showcase

A glimpse of the posts we ship every week.

Every post written, designed and shipped by the Starborn team.

Post written by StarbornPost written by StarbornPost written by StarbornPost written by StarbornPost written by StarbornPost written by StarbornPost written by StarbornPost written by StarbornPost written by StarbornPost written by StarbornPost written by StarbornPost written by StarbornPost written by StarbornPost written by StarbornPost written by StarbornPost written by StarbornPost written by StarbornPost written by StarbornPost written by StarbornPost written by Starborn
Paolo Trivellato, founder of Starborn
Paolo TrivellatoFounder, Starborn
Meet your operator

Built by a team that's done this before.

I'm Paolo. I've spent the last decade getting B2B founders paid from LinkedIn. $4M+ in tracked client revenue across 100+ accounts, and a seat as the in-house LinkedIn coach for two of the largest founder programs in the category.

Starborn is the team I assembled to scale that work. Senior ghostwriters, strategists, and DM operators who have written for some of the most respected personal brands in B2B SaaS, consulting, and venture. Every account is run by people who have already done this at the highest level.

Meet Rajgor
Meet RajgorContent Strategist at Starborn. Your first call is with Meet.
Book your free strategy call →
Who this is for

This isn't for everyone. On purpose.

We turn down more accounts than we take. It keeps the work honest, the team focused, and the case studies real.

Built for
  • B2B SaaS and agency founders doing $30K+ MRR ready to scale
  • Operators who can take 20 to 30 sales calls per month
  • Founders who want to own a category, not chase trends
  • Teams ready to commit to a 90-day window
Not built for
  • Pre-revenue founders without a validated offer
  • B2C, e-commerce, or affiliate plays
  • Founders who want to write every post personally
  • Anyone shopping for viral hacks without a strategy
Final step

Pick a time. Let's talk.

Forty-five minutes with the team. You walk away with a plan whether we work together or not.

  1. 01AuditWe pull apart your current LinkedIn presence, offer, and ICP in real time.
  2. 02MapYou leave with a custom 90-day plan to 20+ qualified calls, written in front of you.
  3. 03DecideAn honest read on whether we're the right partner. No pitch, no pressure.
Book your free strategy call →

Opens Starborn's booking page in a new tab.