agentleFS
duncan-buildroomduncan-buildroom/profit-room-skills/skills/one-lead-magnet/SKILL.md

one-lead-magnet

Skill18 starsChanged 4 months agoMITView on GitHub

Walk a business owner, one question at a time, through building ONE high-value lead magnet and finish with the real thing - a designed PDF plus the launch kit (promo post, image text, DM, five follow-up emails, weekly scorecard). Use when the user says "build my lead magnet", "make a lead magnet", "I need a freebie / opt-in / free guide", "help me get email signups", "comment a word and get a PDF", or "walk me through the lead magnet workbook". It reads your brand profile and any notes, DMs or reviews you point to first, and only asks what is missing.

The file

We split this skill into the rules it gives an agent

11 rules. About 0 are common, 8 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈8

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    If it doesn't exist, offer to run the personalize skill first, or just ask the few basics in Step 1 and keep going.

    line 12
  2. Distinctive:

    If a field this skill needs is blank, ask one quick question and offer to save the answer back to the profile.

    line 13
  3. Distinctive:

    Never invent facts, results, quotes or numbers about the user.

    line 14
  4. Distinctive:

    Use AskUserQuestion when a choice fits (format, title, tool).

    line 24
  5. Distinctive:

    Keep replies short. No lectures.

    line 26
  6. Very specific:

    Read ~/.claude/brand-profile.md if it exists.

    line 30
  7. Distinctive:

    Ask: "Want me to read anything that shows what your customers say?

    line 31
  8. Very specific:

    After each exercise, append the result to ./lead-magnet/answers.md (one ## Exercise N: name block each, plus a ## Status line at the top: …

    line 42
  9. Distinctive:

    Always include: the two-minute self-check up front, a "when it doesn't work / when to get help" section, and a last page with ONE next step …

    line 49
  10. Very specific:

    If the script says NO_CHROME or RENDER_FAILED (common on Windows or without Chrome): tell the member to open lead-magnet.html in any …

    line 57
  11. Distinctive:

    Never put personal health, legal or money details into DMs or emails.

    line 72

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

One Lead Magnet

The conversational version of the Lead Magnet Workbook. You answer plain questions, one at a time. You end up with a finished PDF lead magnet and everything to launch it.

Personalization

This skill works in your brand voice. Before running, load the brand profile at ~/.claude/brand-profile.md.

  • If it doesn't exist, offer to run the personalize skill first, or just ask the few basics in Step 1 and keep going. Never block on it.
  • If a field this skill needs is blank, ask one quick question and offer to save the answer back to the profile.
  • Never invent facts, results, quotes or numbers about the user. Use only what is in the profile, their sources, or their answers.

Profile fields used: Name, What you do, Who you serve, Main pain, What you sell, Transformation, Tone, Words to avoid, Proof, Website, Booking link, Main platforms, brand colors and fonts if present.

How to run it

Rules for every message

  • ONE question at a time, plain language, always with a short example ("Like: ..."). No jargon.
  • Gather first, ask second. Pre-fill whatever the sources already answer, show it, and ask only to confirm or fill gaps.
  • At every decision, give a recommendation and say why in one line. The member can override.
  • Use AskUserQuestion when a choice fits (format, title, tool). Recommended option first.
  • Show progress as "Exercise 3 of 10".
  • Keep replies short. No lectures.

Step 1: Set up and gather

  1. Working folder is ./lead-magnet/ (create it; if you cannot tell which folder the member is in, ask). If ./lead-magnet/answers.md exists, show where they stopped and offer to resume there.
  2. Read ~/.claude/brand-profile.md if it exists. Show a 4-line summary of what you found.
  3. Ask: "Want me to read anything that shows what your customers say? Like a folder of call notes, DMs, reviews, emails, or paste some here." Read what they point to. Never go looking in places they did not name.
  4. Fill in the basics only if still unknown, one at a time: name, business, what you sell, your offer link, who you serve. Offer to save them to the profile at the end.
  5. Show a "here's what I found" summary: buyer questions (with counts), offer, proof, anything else that pre-fills an exercise. Then start.

Step 2: Walk the 10 exercises in order

Open references/exercises.md and follow it exactly for each exercise: pre-fill, questions with examples, recommendation, save.

Part 1 Research: 1 Buyer questions, 2 Problem before your offer, 3 Competitor gap, 4 One-sentence promise. Part 2 Design: 5 Pick the format, 6 Outline and titles. Part 3 Build and Launch: 7 Draft, 8 Proof bank, 9 Your path, 10 Weekly scorecard.

After each exercise, append the result to ./lead-magnet/answers.md (one ## Exercise N: name block each, plus a ## Status line at the top: "Done through exercise N"). Tell the member in one line that it is saved and they can stop and resume any time.

Order: do 1 to 6, then ask the questions of exercises 7 and 8 (stories, regulated field, proof) BEFORE drafting, so the draft already has the member's own stories and proof. Then draft (Step 3), then close out exercises 8, 9 and 10.

Step 3: Build the lead magnet (exercise 7)

  1. Show the approved outline from exercise 6.
  2. Write ONE section at a time into ./lead-magnet/draft.md, using the member's stories, tips, proof and exact words. Short sentences, plain words, the member's tone. Each section ends with one thing the reader does today. Show it and ask for a quick OK or changes before the next. Do not invent facts, results or quotes. Mark approved sections in answers.md.
  3. Always include: the two-minute self-check up front, a "when it doesn't work / when to get help" section, and a last page with ONE next step and ONE link (the offer).
  4. Health, legal and financial niches: put "General information, not advice. This does not replace a [doctor / lawyer / financial adviser]." on the first content page and the last page, keep claims general, no outcome promises, and note the reviewer the member named.
  5. Pay test: "Would you pay $20 for this?" If not, name the section to deepen (depth, not pages) and fix it.

Step 4: Design the PDF

  1. Copy assets/lead-magnet-template.html (in the assets/ folder next to this SKILL.md) to ./lead-magnet/lead-magnet.html. Fill the placeholders, replace the sample sections with the approved draft using the template's components (.section, .do, .proof, ul.check, tables, .lines, .cta). Each .section starts a new page; give very short sections class="section flow" so they share a page. Delete the .note disclaimer div if the niche is not regulated. Before rendering, run grep -c '{{' lead-magnet.html: it must print 0 (no unfilled placeholders).
  2. Set brand colors and fonts in the :root block: from the profile if present, else ask once ("Do you have a brand color? Like: a hex code, or 'dark green'") and fall back to the neutral default. One font family, one accent color.
  3. Render: bash <this skill's folder>/scripts/render-pdf.sh ./lead-magnet/lead-magnet.html ./lead-magnet/<name>.pdf (uses headless Chrome: "/Applications/Google Chrome.app/Contents/MacOS/Google Chrome" --headless --no-pdf-header-footer --print-to-pdf=...). The script also prints the page count.
  4. If the script says NO_CHROME or RENDER_FAILED (common on Windows or without Chrome): tell the member to open lead-magnet.html in any browser, press Ctrl/Cmd+P, choose Save as PDF, turn Headers and footers off, and save.
  5. Check it yourself: convert 2 pages to PNG (pdftoppm -png -r 80 -f 1 -l 2 file.pdf out, then look at the PNGs; if pdftoppm is missing, use the Read tool on the PDF with pages) and look for cut-off text, empty pages, stranded headings, unreadable contrast. Fix and re-render until clean. Tell the member to open it on their phone and confirm they do not have to pinch to read.

Step 5: Launch kit

Write ./lead-magnet/launch-kit.md exactly as specified in references/launch-kit.md: path line, promo post (CTA on the image: "Comment WORD for the [name]"), image text, DM (3 versions, one recommended), five emails, test checklist, weekly scorecard. Show the post, the DM and email 1 for a quick OK; the rest saves to the file.

Step 6: Wrap up

Finish with a short summary: the list of files made (with paths), the PDF page count, and the next step: host the PDF at a link that opens without a login, then deliver it with a comment-keyword DM tool (ManyChat, Blotato or LeadShark; see the course module on delivering with comment-to-DM), connect the email sequence, test from a second account, then post. Offer to save the basics back to the brand profile if they were asked for in Step 1.

Output files (all in ./lead-magnet/)

answers.md (progress and every answer), draft.md (approved text), lead-magnet.html, <name>.pdf, launch-kit.md.

Notes

  • One magnet only. If the promise needs an "and", it is two magnets: keep the stronger half.
  • Pricing for the DM tools changes: tell the member to check each pricing page. This skill writes the text; the tool runs the automation.
  • Never put personal health, legal or money details into DMs or emails.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.