# Viral Linkedin Lead Magnet Post > Use this skill when writing a LinkedIn post that gives away a lead magnet and drives comment-based claims: a giveaway post, "comment X to get it" post, or a rewrite/diagnosis of one that underperformed. Also fires for "write my lead magnet post", "giveaway post for my freebie", "promote my resource on LinkedIn". Produces post copy — hook, pain block, deliverable stack, CTA — calibrated against a 100-post performance dataset of high- vs low-comment posts, plus media direction and a first-comment draft when a sample exists. - Skill: `swan-gtm/viral-linkedin-lead-magnet-post` (Agent Skill, multi-file: 4 files) - Install (CLI): `npx skillmds@latest add swan-gtm/viral-linkedin-lead-magnet-post` - Raw SKILL.md: https://api.skillmd.com/api/skills/swan-gtm/viral-linkedin-lead-magnet-post/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: AI & ML - Author: swan-gtm (https://skillmd.com/u/swan-gtm) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/swan-gtm/viral-linkedin-lead-magnet-post --- # Viral LinkedIn Lead Magnet Post Fires when someone wants a LinkedIn post promoting a free resource, playbook, template pack, or checklist through a "comment WORD to get it" mechanic, or wants an underperforming one diagnosed and rewritten. Produces: the post ready to paste, one media-direction line, a build-scope note, and a first-comment draft when a shareable sample exists. The formula is calibrated against a 100-post performance dataset (top 15 vs bottom 50 by engagement). Read `references/voice-and-structure.md` before writing any copy — it holds the structural rules and the calibration numbers behind them. ## Step 1 — Intake Collect these, extracting silently from what's already given and asking only for the gaps: 1. The magnet — what it is, working name 2. What's inside — the actual deliverables 3. ICP — who it's for, in their words 4. The pain it solves, right now 5. Proof — real numbers or credentials only; never invent these. No proof → use an experience-based or contrarian hook instead 6. Comment word — their preference, or propose one, thematic to the magnet ## Step 2 — Qualify the magnet Check the three rules in `references/titles-and-deliverables.md` (the $100 test, sub-hour consumption, ICP-specific). Fail one → say so plainly and propose the fix before writing, usually by splintering one immediately-usable piece out of a bigger asset. A viral post on a weak magnet burns trust; don't skip this. Rename a vague magnet using the title formulas in the same file. ## Step 3 — Hook angles Generate three hook blocks — hook line, free-claim parenthetical, pivot line — using three different archetypes from `references/hook-angles.md`, matched to the actual proof profile: an angle demanding a metric that doesn't exist is disqualified. Present all three and let the person choose. On "just give me the best one," pick it, state which and why in one line, and go straight to Step 4. For multi-variant runs (up to five), cross each hook archetype with a different narrative frame so variants read as genuinely different posts, not one post re-hooked. ## Step 4 — Build the full post Expand the chosen angle into the five-part structure and every voice rule in `references/voice-and-structure.md`: outcome-first hook, free-claim parenthetical, a pain block that pivots from "I/we" to "you", the deliverable stack (rules in `references/titles-and-deliverables.md`), then CTA + P.S. ## Step 5 — Self-score, then deliver Score the draft against the checklist in `references/voice-and-structure.md` before showing it; fix any failure silently. Deliver the post, one media-direction line (the visual must prove the hook's claim, never decorate it), and a build-scope note that every deliverable bullet is now a promise the resource must contain. If a shareable sample exists, draft the first comment hosting it too — confirm publication rights on real client material, and check any visible date still reads current. ## Rewrites and diagnosis Score the existing post first, name the two or three highest-impact failures in plain language, then rewrite — keep what already works; surgical fixes beat full rewrites when the bones are good. On every returned edit, re-score silently: the comment CTA, the P.S., and the free-claim parenthetical are the usual casualties, and re-audit the deliverable stack against what the fulfilment asset can actually send, since edits routinely add promises it can't keep. Past ~1,600 characters, propose cuts that pay for the addition rather than accepting the bloat. ## What good looks like The single biggest finding from the calibration set: top and flopped posts usually promote equally good resources — packaging is the variable, not the offer. A good post opens on a concrete result, number, or named tool in line one, never a topic or a question. It carries the free-claim parenthetical within the first three lines. It names the reader's exact pain in their own words and pivots to "you" before the deliverable list. The deliverable stack is 4–6 bullets, each concrete enough that the ICP would comment on a public post to get that specific thing. It closes on one comment word and a P.S. that earns its place — a proof pointer, a trust line, or a connection note, never filler. The common failure isn't a weak resource; it's explaining the offer in prose instead of listing it, a vague magnet name, or a fabricated number. Any invented metric, client result, or testimonial disqualifies the post outright. It's ready when every number in it is real and the deliverable stack matches exactly what the fulfilment file will send. ## Hard rules - Never fabricate metrics, client results, or social proof — real numbers or none - One post, one magnet, one CTA — never stack offers - If a proof point sounds invented, ask before using it - Match the person's own spelling convention (UK/US)