Skip to content
Back to search

lead-magnet-factory

MarketplaceWorks with Claude CodeGitHub

A short lead magnet people finish, the comment-to-DM flow and the posts that promote it.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today
Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/mavrin-xv--lead-magnet-factory/badge.svg)](/listings/mavrin-xv--lead-magnet-factory)

HTML

<a href="/listings/mavrin-xv--lead-magnet-factory"><img src="/api/listings/mavrin-xv--lead-magnet-factory/badge.svg" alt="findsafeskills safety rating" /></a>

Install

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

1. Install, then open Claude Code in your project. 2. Paste your input and ask:

   > Here is a voice note from a coffee roaster. We sell a wholesale bean
     subscription to small cafes. Turn this into a lead magnet.

3. The skills run in order: concept card, then `magnets/<slug>/magnet.md`, then the HTML page, then the flow and the posts. You approve each step.

Run the scripts on the invented example first, from the repo root:

python3 skills/magnet-writer/scripts/magnet_lint.py examples/acme-coffee/magnet.md --format checklist --keyword DIALIN
python3 skills/magnet-writer/scripts/md_to_html.py examples/acme-coffee/magnet.md -o /tmp/magnet.html
python3 skills/keyword-delivery/scripts/flow_builder.py examples/acme-coffee/flow.json -o /tmp/flow.md

Sample output:

PASS title: 'Dial in your espresso in 15 minutes: 22 checks for small cafes'
PASS every paragraph is 3 sentences or fewer
PASS no unfilled placeholders
PASS 22 checkbox items
PASS quick win section near the top
PASS Next step has 0 link(s)
PASS keyword DIALIN found in Next step
INFO 917 words, 39 blocks

OK: 0 fail(s), 0 warning(s)

The whole chain for an invented roaster (voice note, 24 survey answers, concepts, the checklist, its HTML and Notion versions, the flow and eight posts) is in [examples/acme-coffee/](examples/acme-coffee/).

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

.claude-plugin/marketplace.json · Manifest · 15 lines
  1. {
  2. "name": "lead-magnet-factory",
  3. "plugins": [
  4. {
  5. "name": "lead-magnet-factory",
  6. "source": "./",
  7. "description": "Turns an article, voice note, research notes or survey answers into a short lead magnet with a quick win (HTML page or Notion), builds and checks the comment-keyword to DM to link delivery flow, and writes give-away posts per platform that promote it.",
  8. "version": "0.1.0"
  9. }
  10. ],
  11. "owner": {
  12. "name": "Artem Mavrin",
  13. "url": "https://github.com/mavrin-xv"
  14. }
  15. }
README.md · README · 141 lines
  1. ![lead-magnet-factory](assets/banner.png)
  2. # lead-magnet-factory
  3. **Turn an article, a voice note, research notes or survey answers into a lead magnet people finish, plus the comment-to-DM delivery flow and the posts that promote it. For creators, consultants and small businesses that sell something after the freebie.**
  4. Most lead magnets are a 30-page PDF nobody opens, sent by a DM flow nobody tested, announced with "comment GUIDE" and nothing else. This pack picks one painful problem, writes a short magnet with a 15-minute win, and checks every step with a script.
  5. ![License: MIT](https://img.shields.io/badge/license-MIT-blue) ![Works with Claude Code](https://img.shields.io/badge/works%20with-Claude%20Code-orange) ![Python 3.8+ stdlib](https://img.shields.io/badge/python-3.8%2B%20stdlib-green)
  6. ```bash
  7. git clone https://github.com/mavrin-xv/lead-magnet-factory && cd lead-magnet-factory
  8. ./install.sh # or: ./install.sh --project (this project only)
  9. ```
  10. <!-- install-notes:start -->
  11. Other ways in: as a Claude Code plugin, run `/plugin marketplace add mavrin-xv/lead-magnet-factory` and then `/plugin install lead-magnet-factory@lead-magnet-factory`. For Codex or another agent that reads `SKILL.md`, copy the folders under `skills/` into its skills directory. On Windows, run `install.sh` from Git Bash or WSL and type `py -3` where these docs say `python3`.
  12. <!-- install-notes:end -->
  13. ## What it solves
  14. | Pain | What the pack does |
  15. |---|---|
  16. | You do not know what to give away | `magnet-concept` drafts 3 to 5 concepts and scores each on five tests (specific, bridge to the offer, quick win, value, AI-proof). A script rejects any concept below 3 |
  17. | You have survey answers but no idea what they say | `survey-to-magnet` clusters the open answers, ranks themes by size, depth and pain words, and pulls the quotes to use in the hook |
  18. | The magnet is long, vague and never finished | `magnet-writer` writes one of six formats with a quick win near the top and a single next step, then `magnet_lint.py` fails it on item counts, long paragraphs, leftover placeholders and missing calls to action |
  19. | Publishing means copying into Notion by hand | One `magnet.md` becomes a standalone HTML page (checkboxes, copy buttons, dark mode, print) and a Notion page through the API |
  20. | The DM flow breaks quietly or gets the account limited | `keyword-delivery` builds the flow from one JSON file and checks keyword, reply length, link placement, button label, disclosure and the 24-hour window, then writes a setup sheet and a test plan |
  21. | "Comment GUIDE" posts get no comments | `magnet-promo` writes a post per platform that gives one real item away first, and `promo_check.py` checks limits, keyword placement and link rules |
  22. ## What's inside
  23. | Skill | What it does | Say |
  24. |---|---|---|
  25. | [magnet-concept](skills/magnet-concept/SKILL.md) | Brief in five lines, 3 to 5 concepts across types and formats, five-test scoring, title, concept card | "lead magnet idea", "what should I give away" |
  26. | [survey-to-magnet](skills/survey-to-magnet/SKILL.md) | Two-question survey, theme clustering of the answers, quotes and a suggested format | "I ran a survey", "what do my customers struggle with" |
  27. | [magnet-writer](skills/magnet-writer/SKILL.md) | Checklist, playbook, swipe file, template, mini-course or hub page, linted, exported to HTML and Notion | "write the lead magnet", "make the checklist" |
  28. | [keyword-delivery](skills/keyword-delivery/SKILL.md) | Comment keyword to DM to link, Telegram start links or form plus email, tagged links, test plan | "comment to DM", "ManyChat flow", "how do I deliver it" |
  29. | [magnet-promo](skills/magnet-promo/SKILL.md) | Give-away posts for X, Threads, LinkedIn, Instagram, TikTok, Telegram and newsletters, plus a launch checklist | "promote my lead magnet", "write the launch posts" |
  30. Scripts (Python standard library, each with `--help`): `score_concepts.py`, `cluster_survey.py`, `magnet_lint.py`, `md_to_html.py`, `md_to_notion.py`, `flow_builder.py`, `promo_check.py`.
  31. ## Quick start
  32. 1. Install, then open Claude Code in your project.
  33. 2. Paste your input and ask:
  34. ```
  35. > Here is a voice note from a coffee roaster. We sell a wholesale bean
  36. subscription to small cafes. Turn this into a lead magnet.
  37. ```
  38. 3. The skills run in order: concept card, then `magnets/<slug>/magnet.md`, then the HTML page, then the flow and the posts. You approve each step.
  39. Run the scripts on the invented example first, from the repo root:
  40. ```bash
  41. python3 skills/magnet-writer/scripts/magnet_lint.py examples/acme-coffee/magnet.md --format checklist --keyword DIALIN
  42. python3 skills/magnet-writer/scripts/md_to_html.py examples/acme-coffee/magnet.md -o /tmp/magnet.html
  43. python3 skills/keyword-delivery/scripts/flow_builder.py examples/acme-coffee/flow.json -o /tmp/flow.md
  44. ```
  45. Sample output:
  46. ```
  47. PASS title: 'Dial in your espresso in 15 minutes: 22 checks for small cafes'
  48. PASS every paragraph is 3 sentences or fewer
  49. PASS no unfilled placeholders
  50. PASS 22 checkbox items
  51. PASS quick win section near the top
  52. PASS Next step has 0 link(s)
  53. PASS keyword DIALIN found in Next step
  54. INFO 917 words, 39 blocks
  55. OK: 0 fail(s), 0 warning(s)
  56. ```
  57. The whole chain for an invented roaster (voice note, 24 survey answers, concepts, the checklist, its HTML and Notion versions, the flow and eight posts) is in [examples/acme-coffee/](examples/acme-coffee/).
  58. ## How it works
  59. ```
  60. input (article / voice note / research / topic) survey CSV
  61. | |
  62. | survey-to-magnet + cluster_survey.py
  63. v v
  64. magnet-concept + score_concepts.py <---- themes and quotes
  65. | concept.md
  66. v
  67. magnet-writer + magnet_lint.py
  68. | magnet.md --> md_to_html.py --> magnet.html
  69. | --> md_to_notion.py --> Notion page (NOTION_TOKEN)
  70. v
  71. keyword-delivery + flow_builder.py --> flow.md (setup sheet + test plan)
  72. v
  73. magnet-promo + promo_check.py --> promo.md (one post per platform)
  74. ```
  75. The model makes the judgement calls: who the reader is, what to give away, every word. The scripts do the counting, so a placeholder left in, a link in a first DM or a caption over the limit is caught every time. Longer walkthrough: [docs/how-it-works.md](docs/how-it-works.md). Markdown the scripts understand: [docs/markdown-conventions.md](docs/markdown-conventions.md).
  76. ## Configure
  77. | What | Where |
  78. |---|---|
  79. | Notion | env `NOTION_TOKEN` (an internal integration token), share the parent page with the integration, then `md_to_notion.py --push --parent <page id>`. Details in `skills/magnet-writer/references/notion-output.md` |
  80. | Page colour | `md_to_html.py --accent "#2f6f4f"` |
  81. | Delivery flow | `flow.json`, start from `templates/flow.example.json` or `flow_builder.py --init` |
  82. | Formats and their structure | `skills/magnet-writer/references/format-structures.md` and the files in `templates/` |
  83. | Platform limits and link rules | the `LIMITS` table at the top of `promo_check.py` |
  84. ## Honest limits
  85. - The scripts were run on the example data only. The Notion push was tested as a dry run; a live push to your workspace was not run by the authors.
  86. - `flow_builder.py` writes a setup sheet. It does not connect to ManyChat, Meta or Telegram, and it cannot see whether your account is allowed to automate DMs.
  87. - Platform rules change. The 24-hour messaging window, one private reply per comment and character limits were current when written; check them before a launch.
  88. - Survey clustering is word overlap, not meaning. "Sour shots" and "under-extracted espresso" can land in different themes. Under 15 answers the themes are a guess, and the script says so.
  89. - The lint checks structure, not quality. A magnet can pass every check and still be dull; the judgement half of the gate in `skills/magnet-writer/references/quality-gate.md` is done by the model and by you.
  90. - No invented proof. The skills refuse numbers that are not in your input or a cited source, which means your first draft may have gaps marked `[NEEDS SOURCE]`.
  91. ## Credits
  92. - Five concept tests, magnet types, interactive formats and their security rules: inspired by the lead-magnet skill in https://github.com/boringmarketer/boring-marketing-skills (MIT).
  93. - Survey to magnet flow (open question, theme ranking, quotes, outline): inspired by `conversion-ops` in https://github.com/ericosiu/ai-marketing-skills (MIT). Rewritten with the standard library, no code copied.
  94. - Format list and gating options: inspired by the lead-magnets skill in https://github.com/coreyhaines31/marketingskills (ideas only).
  95. - Format-to-buyer-stage mapping: inspired by https://www.skills.sh/rampstackco/claude-skills/lead-magnet-design (ideas only).
  96. - Comment-to-DM practice: https://www.planoly.com/blog/instagram-comment-to-dm-best-practices and https://manychat.com/blog/comments-automation (ideas only); platform rules from https://developers.facebook.com/docs/messenger-platform/instagram/features/private-replies/ and https://developers.facebook.com/docs/messenger-platform/policy/policy-overview.
  97. - Reveal-a-problem, sample and one-step magnet types: from Alex Hormozi, *$100M Leads* (ideas only).
  98. - The single open survey question and answer-depth weighting: from Ryan Levesque's Ask Method, https://digitalmarketer.com/ryan-levesque-avatar-campaigns/ (ideas only).
  99. - Repo and skill layout: modelled on https://github.com/anthropics/skills and https://github.com/obra/superpowers (MIT).
  100. ## Contributing
  101. See [CONTRIBUTING.md](CONTRIBUTING.md). Agents working on the repo: [AGENTS.md](AGENTS.md).
  102. <!-- author:start -->
  103. ## Author
  104. Built and maintained by [Artem Mavrin](https://github.com/mavrin-xv). This repo is one of a set of agent skill packs for marketing and operations work. Each one installs on its own; the full list is in [agent-skill-packs](https://github.com/mavrin-xv/agent-skill-packs). If a pack saved you an afternoon, a star helps other people find it.
  105. <!-- author:end -->
  106. ## License
  107. MIT. See [LICENSE](LICENSE).

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.