agent-reach
MUST USE when user wants to 调研/research/搜索/search/查/找/look up anything on the internet — e.g. 全网调研 X / 帮我调研一下 X / 查一下 X / 搜搜 X / 看看大家怎么评价 X / X 上有什么讨论 /…
Generates two genuinely distinct hook variants for the same reel so you can post-test which opening actually wins. Use when the user says "give me two hooks to test", "A/B this hook", "which opening should I use", "hook variants".
$ npx -y skills add Ootto-AI/claude-content-skills --skill ab-hook-tester --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/ab-hook-testerContext preview
The summary Claude sees to decide when to auto-load this skill.
Generates two genuinely distinct hook variants for the same reel so you can post-test which opening actually wins. Use when the user says "give me two hooks to test", "A/B this hook", "which opening should I use", "hook variants".
name: ab-hook-tester description: > Generates two genuinely distinct hook variants for the same reel so you can post-test which opening actually wins. Use when the user says "give me two hooks to test", "A/B this hook", "which opening should I use", "hook variants". user-invokable: true argument-hint: "[the reel idea or script]" license: MIT metadata: author: Ootto version: "1.0.0" category: content
Generates two genuinely distinct hook variants for the same reel so you can post-test which opening actually wins.
You have a reel and two instincts about how to open it.
The script or the idea, plus what you're optimising for (retention, saves, DMs).
Give Claude the input and run this.
You are my hook tester. Reel: [paste script or idea]. I care most about: [retention/saves/DMs]. Give me TWO hooks that use DIFFERENT MECHANISMS — not one hook reworded: - HOOK A: [mechanism, e.g. result-first] — the exact on-screen text + spoken line. - HOOK B: [a different mechanism, e.g. objection-first] — same. Then: 1. HYPOTHESIS: what I learn if A wins, and what I learn if B wins. 2. HOLD CONSTANT: everything that must stay identical (cover, length, caption, posting time). 3. READ: which metric decides it, and at what point (3s retention, not likes). 4. SAMPLE: roughly how many views before the result means anything. If both hooks pull the same lever, say so and replace one.
**Honesty:** One variable. If the cover or caption changes too, it isn't a test — it's two different reels.
**Next:** [reel-scripter](../reel-scripter/SKILL.md) → [cover-thumbnail-brief](../cover-thumbnail-brief/SKILL.md)
---
Built by **[Ootto](https://www.ootto.ai)** — the AI autopilot that connects your tools once and runs the busywork for you, automatically. [Book a demo →](https://www.ootto.ai)
Free, copy-paste Claude skills that run your content growth factory — turn ideas (and your Obsidian memory) into scroll-stopping Instagram reels: hooks, scripts, repurposing, captions, and a 2-week posting calendar.
Repo: Ootto-AI/claude-content-skills
MUST USE when user wants to 调研/research/搜索/search/查/找/look up anything on the internet — e.g. 全网调研 X / 帮我调研一下 X / 查一下 X / 搜搜 X / 看看大家怎么评价 X / X 上有什么讨论 /…
Give Claude a permanent, searchable long-term memory using an Obsidian vault as the store. At the end of a session, summarize what mattered and save it as a…
Reads your reel insights and explains in plain English what to make more of and what to stop — no jargon. Use when the user says "explain my analytics", "what…
Build evidence-led social audience personas from customer interviews, comments, reviews, or CRM notes.
Turns a script into a shoot-ready shot list — every clip, angle and visual you need to film, in order. Use when the user says "what do I need to film", "shot…
Recommends the posting times and weekly cadence that fit your audience's actual activity windows. Use when the user says "when should I post", "best time to…