Hook generator
Generates 20-hook batches for any topic across eight proven patterns — curiosity gap, contrarian, specific number, mistake/warning, transformation, insider secret, question, call-out-the-reader — each tagged with pattern and platform fit, plus a top-3 pick with rationale. Refuses clickbait the content can't cash. Use when the user says "give me hooks for...", "punch up this opening line", "my posts get no traction", or "title options for this video".From its SKILL.md
npx -y skills add alebgl77/claude-inc --skill hook-generatorAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 11 stars11 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
SKILL.md
3.5 KB, 771 tokens by cl100k_base, as published. Nobody here has run it
Hook Generator — Hook Smith
"Scroll-stopping hooks"
When to use
- "Give me hooks for a post about <topic>."
- A finished draft with a flat first line: "punch up this opener."
- Titles and opening lines for a video or Reel, before scripting starts.
- "My content is good but nobody stops scrolling" — retrofit hooks onto existing work.
Workflow
- Ingest the content or topic and list what is genuinely surprising, valuable, or costly in it. A hook may only promise what this step actually finds.
- Note the audience and platform. LinkedIn hooks must survive a 2–3 line fold; X hooks ARE the post; video hooks must be speakable in two seconds.
- Generate 20 hooks covering all eight patterns, at least two per pattern: curiosity gap, contrarian, specific number, mistake/warning, transformation, insider secret, question, call-out-the-reader.
- Tag every hook with [pattern] and [platform fit: LI / X / video / all].
- Cash-check every hook against the content: does the material close this loop, back this claim, justify this number? Kill or rewrite anything that overdraws.
- Rank and pick the top three, each with a one-line rationale naming the audience, the promise, and the platform it fits.
- For the #1 pick, write the follow-through — the one or two lines that come immediately after the hook, so the earned attention isn't wasted.
Output format
## Hooks: <topic>
Audience: <who> | Platform: <where> | Source: <content analysed>
| # | Hook | Pattern | Fits |
|----|------|---------|------|
| 1 | <hook> | curiosity gap | LI / X |
| 2 | <hook> | contrarian | all |
| … | <20 rows total — all 8 patterns, 2+ each> | | |
### Top 3
1. "<hook>" — <pattern> — <why it wins for this audience on this platform>
2. "<hook>" — <pattern> — <rationale>
3. "<hook>" — <pattern> — <rationale>
### Follow-through for #1
<the next 1–2 lines after the hook>
### Killed in cash-check
<n> hooks — e.g. "<hook>" (promised <X>; the content only supports <Y>)
Quality bar
- Exactly 20 hooks; all eight patterns present, at least two each.
- Every hook is specific to THIS content — if swapping the topic still works, the hook dies.
- Zero unpayable promises: every curiosity gap closes inside the content.
- Numbers come from the material, never invented for effect.
- Top-3 rationales name audience and platform, not just "this one is strong".
Example
Invocation: "20 hooks for a post about how we doubled trial conversion by removing the credit-card requirement."
Produces: a 20-row tagged table — specific number: "We deleted one form field. Trial conversion went from 11% to 23%."; contrarian: "Asking for a credit card up front doesn't qualify leads. It disqualifies revenue."; mistake: "The 'best practice' that cost us 400 signups a month." — plus a top-3 with rationales, the follow-through lines for the winner, and two hooks killed in the cash-check for promising "10x" numbers the data doesn't support.
What ships with it
Read from the repository
Just SKILL.md. No reference files, no scripts.
Gives 0 of the 12 instructions most skill authoring skills give in 771 tokens
Counted across 408 of the 422 authors here whose files we hold, read 2026-09-06
- Keep SKILL.md under 500 linesin 118 of 408, across 104 files
- Write skill instructions in imperative formin 44 of 408, across 37 files
- Move detailed reference material into references filesin 41 of 408, across 34 files
- Run init_skill.py to initialize new skillsin 40 of 408, across 33 files
- Draft assertions while test runs are in progressin 32 of 408, across 25 files
- Keep description under 1024 charactersin 31 of 408, across 18 files
- Run package_skill.py once the skill is completein 29 of 408, across 23 files
- Spawn with-skill and baseline runs in the same turnin 28 of 408, across 21 files
- Test added scripts by actually running themin 27 of 408, across 21 files
- Follow the creation steps in orderin 24 of 408, across 17 files
- Generate the eval viewer before evaluating outputs yourselfin 24 of 408, across 17 files
- Write the description in third personin 24 of 408, across 18 files
Said here and by no other author read
- List what is surprising, valuable, or costly in the content
- Note the audience and platform before writing
- Generate 20 hooks covering all eight patterns
- Tag every hook with pattern and platform fit
- Cash-check every hook against the content
- Kill or rewrite hooks that overpromise
Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.