Not in the top 1,000 by GitHub stars
Liamrjohnston/hook-master-skill
A Claude skill suite that critiques and fixes your short-video hook. Paste the link into Claude Code and say: Clone this and add the skill.
History
Review weekly changes to SKILL.md before updating. Dates mark the weeks in which we observed each version; checks apply to the recorded commits.
View weekly changes
four-failures-check/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Instructions only → After Instructions only
287053a825ef → 287053a825ef
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Instructions only
First observation → 287053a825ef
No earlier version was observed for comparison.
hook-formats/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Instructions only → After Instructions only
287053a825ef → 287053a825ef
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Instructions only
First observation → 287053a825ef
No earlier version was observed for comparison.
hook-master/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Network or broad shell → After Network or broad shell
287053a825ef → 287053a825ef
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Network or broad shell
First observation → 287053a825ef
No earlier version was observed for comparison.
lock-in/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Instructions only → After Instructions only
287053a825ef → 287053a825ef
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Instructions only
First observation → 287053a825ef
No earlier version was observed for comparison.
niche-tuner/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Instructions only → After Instructions only
287053a825ef → 287053a825ef
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Instructions only
First observation → 287053a825ef
No earlier version was observed for comparison.
the-council/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Instructions only → After Instructions only
287053a825ef → 287053a825ef
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Instructions only
First observation → 287053a825ef
No earlier version was observed for comparison.
viewer-panel/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Instructions only → After Instructions only
287053a825ef → 287053a825ef
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Instructions only
First observation → 287053a825ef
No earlier version was observed for comparison.
viral-scout/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Instructions only → After Instructions only
287053a825ef → 287053a825ef
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Instructions only
First observation → 287053a825ef
No earlier version was observed for comparison.
visual-match/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Instructions only → After Instructions only
287053a825ef → 287053a825ef
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Instructions only
First observation → 287053a825ef
No earlier version was observed for comparison.
Heat
6.2- Star growth, last 7 daysNot available
- Star growth, last 24 hoursNot available
- Rank movement, last 7 daysNot available
- Commits, last 30 daysNot available
- Accounts asking about it through MCP, last 7 daysNot available
- Hot votes this weekNot available
Heat never changes the order by stars, and sponsorship never touches it. How heat works
Stars over time
30-day change not yet available: not enough history.
Individual skills in this repository
- four-failures-check
four-failures-check - hook-formats
hook-formats - hook-master
hook-master - lock-in
lock-in - niche-tuner
niche-tuner - the-council
the-council - viewer-panel
viewer-panel - viral-scout
viral-scout - visual-match
visual-match
How it works
four-failures-check
When your agent uses it
Checks a video opening for the four ways hooks fail (delay, confusion, irrelevance, disinterest) and points at the exact words or shot that cause each one. Use on any hook draft, script opening or filmed first 3 seconds.
What it loads into context
Frontmatter
59 tokensAlways available to the agent
SKILL.md
742 tokensWhen the skill is activated
Reference documents
417 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- The four failures
- How to run it
- Output (when called on its own)
Files
SKILL.mdcore3,121 bytesreferences/failure-examples.mddoc1,750 bytes
Read SKILL.md
# Four Failures Check
A hook that loses people almost always does one of four things. Check all four, in order, and point at the exact words or frame. Never give a verdict without a quote.
## The four failures
| Failure | What it looks like | The test | The fix |
|---|---|---|---|
| **Delay** | The topic arrives late. Greetings, names, "welcome back", "so", "before we get into it", a slow walk in | Count the words (or seconds) before a stranger knows what the video is about | Cut to the first word that carries the subject. Move the payoff frame to second zero |
| **Confusion** | More than one possible meaning, insider words, a pronoun with nothing behind it, a clever line that needs a second listen | Read it as a complete outsider, not as the expert who wrote it. Could a tired stranger say back what the video is about after one listen? | One meaning. Plain active verbs. Name the thing. Enough words to be clear, not as few as possible |
| **Irrelevance** | It's about the creator, not the viewer. "My journey", "a bit about me", a logo, a title card with the brand | Would the viewer think "this is for me" in the first sentence? | Put the viewer's problem, want or belief first. The creator's story can come later as proof |
| **Disinterest** | Clear and relevant but obvious. Nothing to wonder about. "Gardening tips!!", "Here's how our product works" | Is there a gap between what the viewer believes now and what this video shows? | Add contrast: what they expect versus what's true, the old way versus this way, the fail versus the fix |
Two things that are not failures:View the complete source on GitHubhook-formats
When your agent uses it
Rewrites one video idea as hooks in six formats (Fortune Teller, Experimenter, Teacher, Investigator, Contrarian, and Magician as a visual layer), each built on a real contrast and real proof. Use when someone needs hook options that are genuinely different, not ten versions of the same line.
What it loads into context
Frontmatter
70 tokensAlways available to the agent
SKILL.md
777 tokensWhen the skill is activated
Reference documents
377 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Step 1: Find the contrast
- Step 2: Build the core
- Step 3: The six formats
- Step 4: Pick
- Rules
Files
SKILL.mdcore3,255 bytesreferences/format-bank.mddoc1,511 bytes
Read SKILL.md
# Hook Formats
Ten hooks in one shape are one hook. Six formats give six real choices. Every format needs a contrast underneath it: what the viewer believes or does now, versus what this video shows.
## Step 1: Find the contrast
Write it in one line before any hook: "Viewers think [A]. This video shows [B]."
- Explicit contrast says both sides ("Not more practice. Slower practice.").
- Implied contrast says only B, and the viewer supplies A ("The one setting that fixed my blurry photos").
- If you can't find a contrast the video can prove, the idea is weak. Say so instead of dressing it up.
## Step 2: Build the core
A strong hook usually carries these jobs, not as fixed slots but as meanings: a **subject** (what or who), an **action** (what happened or what to do), an **objective** (the result the viewer wants), the **contrast**, and where the creator has them, **proof** (a number or a visible result) and **time** (how fast). Not every hook needs all six. The first four do most of the work.View the complete source on GitHubhook-master
When your agent uses it
Fixes the first seconds of a short-form video so the right viewers stay. Finds the exact words and first shot that make people swipe, rewrites the hook in six formats, matches it to the visual, makes the next seconds pay it off, and refuses fake stats, trend guesses and view promises. Returns one card you can film today.
What it loads into context
Frontmatter
84 tokensAlways available to the agent
SKILL.md
2,918 tokensWhen the skill is activated
Reference documents
2,728 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- The suite
- Rules that never bend
- Step 0: Get the facts
- Step 1: Pick the route
- Step 2: Run the stages
- The card (your whole reply)
- Self-check before you answer
- Reference files
Files
CREDITS.mddoc3,272 bytesSKILL.mdcore12,025 bytesreferences/edge-cases.mddoc2,260 bytesreferences/worked-example.mddoc5,373 bytes
Read SKILL.md
# Hook Master
Use this when someone wants a better hook, opening line, first three seconds, scroll stopper or intro for a Reel, TikTok, Short or short ad, or says people swipe away early. It runs the whole job and calls eight smaller skills by name.
Your job: make this video's opening clearer and more wanted for the people it is for, using only what the footage and the creator can back up. Then hand back exact words to say, exact on-screen text, the first shot, and the next few seconds. You are not promising views.
## The suite
| Stage | Skill | What it settles |
|---|---|---|
| 1 | `four-failures-check` | Which of the four failures (delay, confusion, irrelevance, disinterest) the current opening commits, pointing at the exact words and shot |
| 2 | `viewer-panel` | Six viewers watch the first 3 seconds. Each says swipe or stay, and why |
| 3 | `the-council` | The panel's verdicts turned into the top 3 changes, ranked |
| 4 | `hook-formats` | The idea rewritten in six formats, each built on a real contrast |View the complete source on GitHublock-in
When your agent uses it
Writes the 1 to 3 sentences straight after the hook so they pay it off and turn attention into trust, and finds the dead stretch that loses people after a good hook. Use when a hook stops the scroll but viewers still leave in the next few seconds.
What it loads into context
Frontmatter
60 tokensAlways available to the agent
SKILL.md
574 tokensWhen the skill is activated
Reference documents
228 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Step 1: Find the gap
- Step 2: Pick one lock-in move
- Step 3: Write it
- Output
- Rules
Files
SKILL.mdcore2,303 bytesreferences/dead-stretches.mddoc963 bytes
Read SKILL.md
# Lock-in
The hook makes a promise. The next 1 to 3 sentences decide whether the viewer believes it. If they get a detour, a backstory or a slow setup, the curiosity turns into "this was bait" and they leave.
## Step 1: Find the gap
Write what the hook promised, then what actually comes next in the video, with the time. Flag any stretch that doesn't serve the promise: an intro, a story about the creator, "before we start", a long setup. Say exactly what to cut or move.
## Step 2: Pick one lock-in move
| Move | When | Example shape |
|---|---|---|
| **Proof right away** | The hook made a claim | Show the result, the number (theirs), the screen |
| **Picture the result** | The viewer wants a change | "By the end you'll [see / have] [result]" said over the result |View the complete source on GitHubniche-tuner
When your agent uses it
Tunes a hook to one niche and one viewer, using the words that audience actually uses, stakes the right size, proof they trust and lines the platform and regulator allow. Use when a hook is clear but sounds generic, or when it is for a regulated or local business.
What it loads into context
Frontmatter
63 tokensAlways available to the agent
SKILL.md
728 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Step 1: Name the viewer in one line
- Step 2: Steal their words, not your words
- Step 3: Size the stakes
- Step 4: Use proof this niche trusts
- Step 5: Check the rules of the niche
- Output
Files
SKILL.mdcore2,962 bytes
Read SKILL.md
# Niche Tuner
A hook that could fit any niche fits none. Tune it to the exact viewer.
## Step 1: Name the viewer in one line
"[Who], who [situation], and wants [result] / fears [problem]."
Example: "New runners in Bristol who want to finish a 5k but think their knees can't take it."
## Step 2: Steal their words, not your words
Look for the viewer's own phrases in what the creator gave you: comments, DMs, questions they get asked, what people say before they book. Those words go in the hook. Only words that appear in the brief count as theirs. A phrase you think they would use is your guess: you may use it, but never put it in quotes or say commenters say it.
| Creator word | Viewer word |View the complete source on GitHubthe-council
When your agent uses it
Turns a set of hook verdicts (from the viewer panel, the four failures check or user feedback) into the top 3 changes, ranked, with exact before and after words. Use when you have many notes on a hook and need to decide what to change first.
What it loads into context
Frontmatter
64 tokensAlways available to the agent
SKILL.md
399 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- How to rank
- Output
- Rules
Files
SKILL.mdcore1,635 bytes
Read SKILL.md
# The Council
Notes pile up. The Council decides. It takes the panel's verdicts and the failures found, and returns three changes, ranked, each with the exact before and after.
## How to rank
Score each possible change on three things, then rank:
1. **Who it keeps.** How many of the viewers who matter for this video it turns from swipe to stay. The two "who matter most" viewers count double.
2. **Can the footage pay it off.** A change that needs footage or proof the creator doesn't have drops to the bottom or out.
3. **Cost.** Reorder or cut existing footage is cheapest. New text is cheap. A reshoot is dearer. Say which it is.
Fix in this order when it's a tie: delay, then confusion, then irrelevance, then disinterest. A clear, viewer-first opening with weak curiosity still beats a curious one nobody understands.
View the complete source on GitHubviewer-panel
When your agent uses it
Six different viewers watch the first 3 seconds of a video and each says swipe or stay, and why. Use to pressure-test a hook from the point of view of the doom-scroller, the sceptic, the beginner, the niche expert, the buyer and the rival creator.
What it loads into context
Frontmatter
64 tokensAlways available to the agent
SKILL.md
667 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- The six viewers
- How to run it
- Rules
Files
SKILL.mdcore2,802 bytes
Read SKILL.md
# Viewer Panel
A hook is judged by strangers in a few seconds. Simulate six of them. Each one watches only what is actually in the first 3 seconds: the first shot, the on-screen text and the spoken words. They know nothing else.
This is a thinking tool, not data. Never present the panel as a survey, a prediction or a percentage of real viewers.
## The six viewers
| Viewer | Who they are | What makes them stay | What makes them swipe |
|---|---|---|---|
| **Doom-scroller** | Thumb moving, half watching, sound maybe off | A clear picture of something surprising or wanted, readable text | Talking head with nothing happening, a greeting, a logo |
| **Sceptic** | Has seen a thousand hooks, smells clickbait | A specific, checkable claim with proof on screen | "Nobody tells you", "secret", round stats with no source, hype |
| **Beginner** | New to the topic, doesn't know the words | Their own problem in their own words | Jargon, insider terms, assumes knowledge |
| **Niche expert** | Knows the topic well | A real detail or a claim that's new to them | The obvious advice they already know |View the complete source on GitHubviral-scout
When your agent uses it
Finds the pattern behind hooks that worked in a niche, never the wording. Works honestly without web access by asking the user to paste examples. Use when someone asks what's trending, what's going viral, or wants to learn from other creators' hooks.
What it loads into context
Frontmatter
57 tokensAlways available to the agent
SKILL.md
562 tokensWhen the skill is activated
Reference documents
236 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- First, say what you can see
- The honest route
- Break down the pattern, not the words
- Use it without copying
- While you wait for examples
- Never
Files
SKILL.mdcore2,505 bytesreferences/pattern-sheet.mddoc833 bytes
Read SKILL.md
# Viral Scout
People ask "what's going viral this week?" The honest answer depends on what you can actually see.
## First, say what you can see
- **No web or platform access (the usual case):** say so in one plain line. "I can't see this week's feeds, so I won't guess what's trending." Then offer the useful route below. Never list "trending hooks" from memory as if they were current.
- **Web access:** you can search, but view counts on a page are a snapshot and creators' own numbers are self-reported. Say where each example came from and when.
## The honest route
Ask for 3 to 5 examples, pasted or linked:
- The first line spoken, the on-screen text, what's in the first shot.
- Rough views, and that creator's usual views if they know them. A video that beats its own creator's normal is more useful than a big account's ordinary post.View the complete source on GitHubvisual-match
When your agent uses it
Makes the first-second visual, the on-screen text and the spoken words say the same thing, using footage the creator already has. Use when a hook is written but the opening shot is a talking head, a logo or an unrelated frame, or when planning a video from its key visual.
What it loads into context
Frontmatter
65 tokensAlways available to the agent
SKILL.md
653 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Three channels, one message
- Step 1: Find the key frame
- Step 2: Pick a visual hook type
- Step 3: Write the text
- Step 4: Check the match
- Output
Files
SKILL.mdcore2,664 bytes
Read SKILL.md
# Visual Match
Most people decide with their eyes before they hear a word, and many watch with the sound off. If the picture says "man on a sofa" and the words say "this garden took one weekend", the picture wins and they swipe.
## Three channels, one message
| Channel | Job | Common miss |
|---|---|---|
| Visual (0 to 1 s) | Show the subject, ideally the most surprising or wanted frame | Talking head at a desk, logo, black screen, slow walk in |
| On-screen text | Make the topic clear with sound off, 3 to 7 words, adds to the speech | Repeats the speech word for word, or a vague label ("Tips!!") |
| Spoken | The hook line | Starts with a greeting or a name |
They must point at the same subject. If they don't, change the cheapest one.
View the complete source on GitHubSkills included
- four-failures-check
- hook-formats
- hook-master
- lock-in
- niche-tuner
- the-council
- viewer-panel
- viral-scout
- visual-match
Repository facts
- License
- MIT
- Created
- Oct 25, 2026
- Last pushed
- Oct 25, 2026
- Forks
- 4
- Open issues
- 0
- Project activity
- Active
Before you install
Network or broad shellAutomated signals, not a guarantee
Checked Nov 6, 2026
- No broad tool permissions
- No dangerous shell commands
- 1 network referencein 1 file: hook-master/CREDITS.md:39
- No hidden or obfuscated text
- No instruction overrides
- No executable scripts
- No access to secrets
- No requests to vote in Hot
Show the signal with its line
hook-master/CREDITS.md
- Network accessline 39
Ideas only, no text: dontbesilent2025/dbskill (`dbs-hook`, `dbs-title-cover-intro`), licensed CC BY-NC 4.0. We used it as the comparison in our blind test and took two ideas from it in our own words: check the content before fixing the hook
How the signals fit together
Facts from a static read of the files at the pinned commit. They describe what the code does, not why.
Some files could not be read; the facts below still hold.
Signal confidence: 1 fact
- Network access in prose of documentationhook-master/CREDITS.md:39Low confidence
Works with
| Agent | Status | Evidence | Install to |
|---|---|---|---|
| Claude Code | Compatible | Standard SKILL.md four-failures-check/SKILL.md | .claude/skills/<skill-name>/ |
| Codex | Compatible | Standard SKILL.md four-failures-check/SKILL.md | .agents/skills/<skill-name>/ |
| Cursor | Compatible | Standard SKILL.md four-failures-check/SKILL.md | .cursor/skills/<skill-name>/ |
| Gemini CLI | Compatible | Standard SKILL.md four-failures-check/SKILL.md | .gemini/skills/<skill-name>/ |
| GitHub Copilot | Compatible | Standard SKILL.md four-failures-check/SKILL.md | .github/skills/<skill-name>/ |
| OpenCode | Compatible | Standard SKILL.md four-failures-check/SKILL.md | .opencode/skills/<skill-name>/ |
- Native The repository ships a manifest, folder or frontmatter field for this agent.
- Compatible A standard SKILL.md that this agent can read.
- Unknown No evidence either way.
Derived automatically from paths, manifests and frontmatter. Not tested on each agent.