
AI skills for human-sounding copy and ethical hooks, with a zero-dependency slop scanner.
AI skills for human-sounding copy and ethical hooks, with a zero-dependency slop scanner.
·
Skills that help an AI assistant write copy people actually want to read, and design flows people actually want to finish. It also ships a small scanner that tells you when a text sounds like a machine wrote it.
The test I apply to everything here: if the reader saw your design notes, would they still trust the page?
1. Install the skills
Claude Code plugin (easiest):
/plugin marketplace add Vumb-VibeCoder/hook-everything
/plugin install hook-everything@hook-everythingOr copy the files by hand:
git clone https://github.com/Vumb-VibeCoder/hook-everything.git
cd hook-everything
# Claude Code
mkdir -p .claude/skills .claude/tools
cp -r skills/* .claude/skills/
cp -r tools/* .claude/tools/For Cursor, Windsurf, Codex or other agents, copy AGENTS.md into your project root (or append it to .cursorrules / CLAUDE.md) and keep skills/ and tools/ in the workspace.
2. Ask your assistant
Rewrite this product announcement so it sounds human, then check it with the scanner.3. Check any text yourself
python3 tools/slop_scan.py draft.txt --fix-hintsYou get a score from 0 to 100 and a list of what to rewrite. Aim for 15 or lower. Needs Python 3.10+ and nothing else.
| I want to... | Skills (in order) |
|---|---|
| Humanize or clean up a text in any language | anti-slop → self-review → value-check |
| Write a short email, release note or headline | writing-hooks → anti-slop → self-review → value-check |
| Write an essay, article or newsletter | audience-calibration → trust-architecture → hook-compounding → writing-hooks → anti-slop → self-review → value-check |
| Build a landing page, ad or outreach sequence | market-intelligence → audience-calibration → trust-architecture → hook-compounding → marketing-hooks → visual-ux-hooks → anti-slop → self-review → value-check |
| Design onboarding or a product flow | psychology-map → trust-architecture → visual-ux-hooks → hook-compounding → product-hooks → value-check |
| Find fresh angles and ideas | market-intelligence → creative-ideation → hook-compounding → anti-slop → self-review → value-check |
Short edits use the Lite Lane (2 to 3 skills). Campaigns and new products use the Full Lane. Details are in AGENTS.md.
| Skill | What it does |
|---|---|
psychology-map | 30 behavior mechanisms, each rated by evidence strength |
trust-architecture | Builds trust before asking for anything |
market-intelligence | Reads live reviews and competitor gaps, and drops forced trend-jacking |
audience-calibration | Matches speed, tone and proof to the reader |
creative-ideation | Throws out the first obvious ideas and combines unrelated ones |
hook-compounding | Layers 3 to 4 mechanisms into one hook |
visual-ux-hooks | Layout, contrast, choice limits and response timing |
marketing-hooks | Conversion copy without fake urgency or inflated claims |
writing-hooks | Openers and mid-article retention |
product-hooks | Onboarding, habits and humane offboarding |
anti-slop | 85 rules for natural, non-hype writing |
self-review | Rereads a draft as a native reader in any language and rewrites it if it still sounds machine-made |
value-check | 13-point ethics and quality gate before you ship |
# Score a file and get rewrite hints
python3 tools/slop_scan.py draft.txt --fix-hints
# Scan a doc that quotes banned phrases on purpose
python3 tools/slop_scan.py skills/anti-slop/SKILL.md --strip-markdown
# Fail a CI job when the score is above 15
python3 tools/slop_scan.py README.md --max-score 15
# Report invisible Unicode (zero-width, U+202F, bidi, tag block, lookalikes)
python3 tools/slop_scan.py draft.txt --hidden
# Write a cleaned copy, or fail a CI job if any hidden character is found
python3 tools/slop_scan.py draft.txt --clean draft.clean.txt
python3 tools/slop_scan.py draft.txt --fail-on-hidden
# JSON output, or run the built-in tests
python3 tools/slop_scan.py draft.txt --json
python3 tools/slop_scan.py --selftestClaude does, but not with characters. Anthropic says its watermark biases word choice using a secret key, adds no hidden characters, and can only be checked by someone holding the key. So no local tool can detect or strip it, this scanner included. ChatGPT text sometimes carries odd spaces such as U+202F, but those also come from copy-paste and are weak evidence. --hidden finds and removes invisible characters from any source. Sources and caveats are in research/WATERMARKS.md.
The 15-point threshold is a working heuristic for dense clusters of AI and marketing tropes. It does not prove who wrote a text.
hook-everything/
├── AGENTS.md # Routing rules for agents
├── SKILL.md # Root skill manifest
├── skills/ # The 13 skills
├── tools/ # slop_scan.py
├── tests/ # unittest suite
├── research/ # Sources, evidence ratings, watermark notes
└── examples/ # Before/after rewritesRun the checks before a pull request:
python3 tools/slop_scan.py --selftest
python3 -m unittest discover -s testsSee CHANGELOG.md for version history. Licensed under MIT.
This repo does not document an install step we could read. Check the README on GitHub.
Paste this prompt into your agent. It reads this page and does the setup for you.
Read https://aiagentslisting.com/mcp/hook-everything to learn what the "Hook Everything" MCP server does and how to install it. Add it to my coding agent's MCP configuration as documented on that page, then confirm the server connects and list the tools it exposes.Agents can also browse this directory over MCP at https://aiagentslisting.com/api/mcp
This server doesn't publish a machine-readable tool manifest, and it doesn't expose a public endpoint we could ask. Connect the server locally to see what it exposes.
We could not run this server: it ships no npx package for our sandbox and no public endpoint. The four checks that need a live server are left out, and the grade tops out at B until one exists.
Maintain this server? Claiming proves you own the listing and earns the last 4 points.
Claim this listingPaste this line near the top of your repository's README. The badge always renders the current score.
[](https://aiagentslisting.com/mcp/hook-everything)Nothing comparable is listed yet.
No tagged releases on record.
Connect to the AI Agents Listing MCP server at https://aiagentslisting.com/api/mcp using the HTTP transport and complete the sign-in when it asks. Then call the submit_review tool for the listing "hook-everything" with a rating from 1 to 5 and a short note on what worked and what did not.One email a week. New agents, MCP servers and skills, and what is actually getting traction.