linkedin-hooks
Skill1 starsChanged 23 days agoMITView on GitHub
Write and score opening lines for a LinkedIn post before the full post gets written. A small local tool scores each hook 0 to 100 on six checks (fits before "see more" on a phone, concrete, tension, a person in it, plain words, no engagement bait) and the skill explains what to fix. Use it when someone says "give me hooks", "better first line", "which opening is best", "score these hooks", "my posts get no clicks on see more", or pastes a few first lines to compare. Reads the active business from founder-context so hooks fit its topics and avoid its banned words. Hands the winner to linkedin-post.
1 other project uses a version of this.
Adopted by
Most recently taken up by
- matteotitta/claude-marketing-linkedintheir own version, first seen Oct 25, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- matteotitta/claude-marketing-linkedin★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository15 other files it gives its agents.
- founder-contextSkill
- humanizeSkill
- linkedin-auditSkill
- linkedin-carouselSkill
- linkedin-commentSkill
- linkedin-dmSkill
- linkedin-inboxSkill
- linkedin-planSkill
- linkedin-postSkill
- linkedin-profileSkill
- linkedin-replySkill
- linkedin-repurposeSkill
- linkedin-viralSkill
- talking-head-cardsSkill
- transcribeSkill
The file
We split this skill into the rules it gives an agent
12 rules. About 0 are common, 11 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈11
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 31
If the user gave only a topic, ask what happened, with what number, for whom.
- Very specific: line 36
Use different patterns from references/patterns.md (number, before-after, mistake, scene, said-no, claim, question, overheard).
- Distinctive: line 41
Write each hook as the founder would say it out loud, in the business's tone words.
- Distinctive: line 42
Keep each under about 140 characters.
- Distinctive: line 44
Use only what the user said or the business file holds.
- Distinctive: line 62
no bait (10): no "Agree?", "comment below" and the like.
- Distinctive: line 82
If the top hook scores under 70, say what is missing (usually a real number or a real moment) and ask for it rather than polishing weak …
- Distinctive: line 85
If the user pasted their own hooks, score theirs first and show them, then offer at most 3 improved versions that keep their words where …
- Distinctive: line 90
When you stop to ask for facts instead of writing hooks, still end with the handoff line, so the user knows linkedin-post comes after.
- Distinctive: line 103
Never write a hook that claims something the founder did not do or say.
- Distinctive: line 104
Never use engagement bait to lift a score.
- Distinctive: line 105
Never post anything. The user posts.
≈ “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-hooks
Line 1 decides whether anyone taps "see more". This skill writes options, scores them with a local tool, and says why the winner wins. It does not write the full post.
Tool paths: <this skill> is the folder this SKILL.md is in (the "base directory" you
were given). Sibling skills sit next to it, so <this skill>/../linkedin-hooks/ is the
hooks skill. Run tools with full paths built from that folder, not from the current
working directory, or they will look missing.
1. Business and idea
Find the business the same way every skill in this repo does: a named business, else
founder/context.md, else ~/.founder/active and
~/.founder/businesses/<slug>.md. If none exist, ask once for topics, tone words,
words to avoid, and proof, and offer the founder-context skill. Never invent proof.
Name the business in your first line. Then get the one true thing the post is about. If the user gave only a topic, ask what happened, with what number, for whom. A hook without a fact under it is a promise the post cannot keep.
2. Write 6 to 8 options
Use different patterns from references/patterns.md (number, before-after, mistake,
scene, said-no, claim, question, overheard). Only use a pattern the facts support:
no "overheard" without real words, no "number" without a real number. When a number is
missing, write {{your number}} and keep going.
Write each hook as the founder would say it out loud, in the business's tone words. Keep each under about 140 characters. Do not picture scenes, add feelings, or work out new numbers ("9 on-time weeks") to make a hook stronger; those are made up even when flagged. Use only what the user said or the business file holds. Habits and opinions count too: "I count every late drop" or "I think parades are worse than rain" are made up unless the user said them. Pick the top scorer unless you say why another one is better.
3. Score them
Save the options one per line to a temp file and run:
python3 <this skill>/scripts/hookscore.py hooks.txt --topics "<topic 1>,<topic 2>" --avoid "<words to avoid>"The tool checks six things (weights in brackets):
- fold (15): fits before "see more" on mobile, about 140 characters.
- concrete (25): a number, a time, or a name.
- tension (20): something changed, cost, broke, or was refused.
- person (10): "I", "we" or "you" is in it.
- plain (20): no stock phrases, no vague words, no business words to avoid.
- no bait (10): no "Agree?", "comment below" and the like.
The tool is a heuristic written for this repo. It ranks options. It does not predict reach, and a 95 can still be wrong for this audience. If Python is not available, score by hand on the same six checks and say so.
4. Show the result
HOOKS for <business>, idea: <one line>
1. 92 <hook> pattern: number
2. 85 <hook> pattern: before-after
3. 78 <hook> pattern: mistake
(lower ones in one line each, with the main note)
Pick: #1. <one sentence on why, in plain words>
Fix before posting: <each {{placeholder}}>, or "nothing"If the top hook scores under 70, say what is missing (usually a real number or a real moment) and ask for it rather than polishing weak lines.
If the user pasted their own hooks, score theirs first and show them, then offer at most 3 improved versions that keep their words where they can. If one of their hooks names a fact that is not in the business file (a family member, a customer, a number), ask them to confirm it is real and OK to share before you pick it.
When you stop to ask for facts instead of writing hooks, still end with the handoff line, so the user knows linkedin-post comes after.
5. Save and hand off
Append the scored list to ~/.founder/social/linkedin/<slug>/hooks.md under a heading like
## 2026-10-23, idea: <one line> (a comma, not a dash), so linkedin-audit can later see
which patterns worked.
Next: "Say 'write it' and linkedin-post will draft the post on hook #1."
Never
- Never write a hook that claims something the founder did not do or say.
- Never use engagement bait to lift a score. The tool penalizes it for a reason.
- Never post anything. The user posts.
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.

