linkedin-lead-generation
Skill1 starsChanged 11 days agoMITView on GitHub
Turns LinkedIn attention into booked calls and paying clients for founders, consultants, and creators selling their own services. Finds the broken step in the content-to-client path, fixes the one-line offer, scripts inbound plays and DM replies, and plans capped, personalized warm outreach. Use when someone asks how to get leads, clients, or calls from LinkedIn, how to reply to DMs, or how to do outreach. Refuses mass automation, scraping, and templated DM spam.
In this repository5 other files it gives its agents.
The file
We split this skill into the rules it gives an agent
8 rules. About 0 are common, 6 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈6
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 31
Every lead follows the same route.
- Distinctive: line 48
Never optimize step 4 when step 1 is broken.
- Distinctive: line 53
Before any outreach, the user must be able to finish this in one line:
- Distinctive: line 58
If it takes a paragraph, prospects will not repeat it to their network and DMs will stall.
- Distinctive: line 104
Never: pitch right after a connection is accepted, send a voice note to a stranger, or follow up more than once on an unanswered thread.
- Very specific: line 109
Only once inbound is running.
- Very specific: line 120
Warm up a dormant account first: If the user has been inactive for months, follow the warm-up protocol in references/platform-reality.md …
- Distinctive: line 140
Ask three questions in DM before booking: what they are trying to fix, what they have already tried, and their timeline.
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
LinkedIn Lead Generation
Convert LinkedIn attention into conversations, and conversations into clients.
Hard boundary: this skill does personalized, human outreach at human volume. It does not build automated connection blasts, scraped lead lists, or templated DM spam. Those get accounts restricted and burn reputation permanently. If the user asks for mass automation, say no once, explain why in a sentence, and offer the version that works.
Pick the depth
- Quick: the user needs one thing ("how do I reply to this DM", "fix my offer
line"). Do that step only, using
references/dm-scripts.mdfor messages, and name the broken step if the numbers they gave point to one. - Full (default for "help me get leads"): the path diagnosis, then Steps 1 to 6.
The path
Every lead follows the same route. Find the broken step before optimizing anything else.
Content seen -> profile visited -> offer understood -> conversation started -> call booked -> clientDiagnose with the user's own numbers:
| Symptom | Broken step | Fix with |
| Posts get no views | Content | linkedin-content-strategy |
| Views but no profile visits | Hook and identity in the post | linkedin-content-strategy |
| Profile visits but no DMs | Profile conversion | linkedin-profile-audit |
| DMs but no calls | Offer clarity and DM handling | this skill, Steps 1 and 3 |
| Calls but no clients | Offer, price, or qualification | this skill, Step 5 |
Never optimize step 4 when step 1 is broken. Name the broken step, with the number that proves it, before giving any tactic.
Step 1: Make the offer legible
Before any outreach, the user must be able to finish this in one line:
"I help [specific audience] achieve [specific outcome] in [timeframe], usually via [mechanism]."
If it takes a paragraph, prospects will not repeat it to their network and DMs will stall. Fix this first: it is the most common real blocker.
Then place it where traffic already lands: the headline, the About CTA, Featured slot 1, and the pinned post.
Step 2: Inbound plays (run these before outreach)
Inbound converts better and costs no reputation.
Comment-to-deliver, used sparingly. End a high-value post with "Comment [WORD] and I'll send it", reply to each comment, then DM the asset. It creates real conversations with real consent. But LinkedIn reduces the reach of posts that ask for engagement mainly to boost reach, so use it rarely, only for an asset worth the ask, and also put the link in the first comment for anyone who does not want to comment.
A lead magnet that proves competence. A checklist, teardown template, or calculator that solves one narrow problem end to end. Ten useful pages beat a 50-page ebook nobody opens.
Profile-to-call path. Featured slot 1 is the booking link. The About ends with the same link. Both use the same words as the offer line.
Free teardown play. Publicly audit a real example (with permission, or a public brand), post the findings, and let readers ask for theirs. It converts better than any cold DM because the work is visible before the ask.
Step 3: DM handling
Most founders lose leads here by pitching too early. Messages follow
references/voice-rules.md, and the templates are in references/dm-scripts.md,
written to be edited, never sent as-is.
Sequence for an inbound DM:
- Acknowledge and give. Deliver the promised thing with no strings.
- One question. About their situation, not your service: "What made you grab this? What are you trying to fix?"
- Diagnose in the thread. Give one genuinely useful answer for free. This is the sale.
- Soft bridge, once. "This is exactly what I do with clients. Want me to walk you through how I'd fix it on a call?"
- If they say no or go quiet, stay useful. One follow-up, 3–4 weeks later, with something relevant. Then stop.
Never: pitch right after a connection is accepted, send a voice note to a stranger, or follow up more than once on an unanswered thread.
Step 4: Warm outreach (capped volume)
Only once inbound is running. Every figure here comes from
references/platform-reality.md; check it there before quoting it.
- Weekly budget first. LinkedIn's invitation limit is reported at about 100 a week. Send 10–15 personalized requests a weekday. 20 a day is the ceiling, not a target: it spends the whole weekly budget in five days.
- Daily action ceiling. Invitations, DMs, profile views, likes, comments, and endorsements all count toward a reported ceiling near 120 a day. Stay well below it.
- Watch the acceptance rate. Above 30% is healthy. Between 15% and 30%, slow down and tighten targeting. Below 15%, stop and fix targeting before sending more.
- Warm up a dormant account first. If the user has been inactive for months,
follow the warm-up protocol in
references/platform-reality.mdbefore any invitations. - Target only active people. People who posted or commented in the last 30 days (Sales Navigator can filter for this; otherwise check their activity by hand). Requests to dormant profiles burn a capped weekly budget for nothing.
- Targeting sources: people who engaged with the user's posts, people who viewed the profile, second-degree connections of happy clients, and members of the same communities. Never a scraped or bought list.
- Note or no note: send a note only with a real, specific reason (they commented, you share a client, you read their post). Otherwise send it blank: generic notes lower acceptance.
- Keep messages to 3–4 sentences, sent during the recipient's business hours.
- The first message is never a pitch. Reference the actual reason you connected.
Track acceptance rate, reply rate, and calls booked per 100 requests. If the reply rate is under 10%, the targeting is wrong, not the script.
Step 5: Qualify before the call
Ask three questions in DM before booking: what they are trying to fix, what they have already tried, and their timeline. It protects the calendar and raises the close rate, because the call starts mid-conversation.
Disqualify openly when it is not a fit, and refer them elsewhere. It costs one lead and buys a lasting reputation.
Step 6: Deliver
- Which step of the path is broken, with the number that proves it
- The one-line offer, rewritten
- Two inbound plays to run in the next 14 days, scripted
- The DM sequence adapted to their offer
- An outreach plan with daily and weekly caps and targeting sources, only if inbound is live
- Metrics: profile views, DMs started, calls booked, close rate, with today's baseline (or "not provided")
Anti-patterns
- Pitching in the connection request. The fastest way to be ignored and reported.
- Automation tools that send on the user's behalf. Against LinkedIn's terms and a restriction risk.
- Buying lead lists or scraping at volume.
- "Quick question" openers and fake personalization.
- Booking calls with unqualified people to feel busy.
- Measuring followers instead of conversations. Report conversations.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

