--- name: Pick Your Platform category: content works_with: [claude, chatgpt, gemini] difficulty: plug-and-play version: 1.0 source: adapted (viral "7 social media prompts" pack — one question rebuilt, six prompts rejected as covered) tested: no --- # Pick Your Platform ## What it does Decides where your content should live before you make any — based on where the people you want are actually spending time, what format you can genuinely keep producing, and how many platforms one person can sustain (fewer than you think). The decision most content advice skips straight past on the way to a posting schedule. ## When to use it - You're posting to four platforms and none of them is working - Starting out and everyone says "be everywhere" - You've been consistent for months on a platform your audience doesn't use ## The skill ``` Act as a practical content planner. HARD RULES: no claims about what any platform's algorithm rewards — that's unknowable from outside and changes constantly; never predict follower growth, reach, or timelines; never invent audience statistics — if I don't know where my people are, tell me how to find out rather than guessing. Size everything to the hours I actually have. Bias toward fewer platforms; if I insist on five, say plainly what that costs. Who I want to reach: [SPECIFIC PEOPLE — not "everyone interested in X"] Where I know or suspect they already spend time: [PLATFORMS, GROUPS, NEWSLETTERS, PODCASTS — or "no idea"] What I can actually make and keep making: [WRITING / TALKING TO CAMERA / SHORT VIDEO / AUDIO / IMAGES — honestly, given my skills and how I feel on camera] Hours per week, really: [NUMBER] What I want this to do: [FIND CLIENTS / BUILD AN AUDIENCE / SELL A THING / SHARE WORK] Where I already have any footing: [EXISTING ACCOUNTS, LIST, NETWORK] Work it out: 1. WHERE THEY ARE: from what I told you, where my specific people most plausibly gather — and if I said "no idea", the three cheapest ways to find out this week (ask five of them, look where competitors' comments come from, check where the questions in my niche get asked). 2. THE FORMAT MATCH: which of those places suits what I can sustainably make. A platform I'd have to become a different person to serve is the wrong platform, however big it is. 3. THE ONE: the single platform to commit to first, with the honest reasoning — and the one I'm tempted by that I should leave alone for now, and why. 4. THE ROLE OF ANY SECOND: if my hours genuinely allow it, what a second platform would do that the first can't (a home for the long version, a place to be found by search, where clients actually message). If they don't, say so. 5. THE REPURPOSE PATH: how one piece made for the main platform gets reshaped for anywhere else — so any second presence costs minutes, not a second production line. 6. THE REVIEW DATE: when to judge this decision, what I'd look at (my own numbers, not benchmarks), and what result would mean it's time to switch rather than push harder. ``` ## Example output [TO FILL AFTER TESTING] ## Tweaks - One platform done consistently beats four done sporadically — and consistency is a function of hours, not motivation - Pick the format you can make on a bad week; the good weeks take care of themselves - Where your people already ask questions is more useful than where the biggest audience is - Pair with Find Your Angle before this, and Shorts Batch Week or Daily Three-Post Engine after it --- --- name: YouTube Hook Generator category: content works_with: [claude, chatgpt, gemini] difficulty: plug-and-play version: 1.0 source: original tested: no --- # YouTube Hook Generator ## What it does Generates 10 opening-line hooks for a video topic, each under 12 words, built to stop the scroll in the first 2 seconds. ## When to use it - You have a video topic but the first line feels flat - Your retention graph shows a cliff in the first 10 seconds - You're batch-scripting multiple videos and need hooks fast ## The skill ``` You write YouTube hooks. Rules: - Max 12 words per hook - No greetings, no "in this video", no channel intro - Start mid-thought, as if the viewer walked into a conversation - Each hook uses a different mechanism: contradiction, specific number, open loop, confession, before/after, direct challenge, pattern break, cost of inaction, unexpected comparison, one-line story - Plain spoken language. Nothing that sounds like a marketer wrote it. My video topic: [TOPIC] My angle / what actually happened: [1-2 SENTENCES] Give me 10 hooks, labeled by mechanism. Then mark the one you'd bet on for retention and say why in one sentence. ``` ## Example output [TO FILL AFTER TESTING] ## Tweaks - Add "my audience is [X]" for niche-specific language - Change max words to 8 for Shorts - Ask for title + thumbnail-text variants of the winning hook to keep them connected but different --- --- name: Plain Email Rewriter category: business works_with: [claude, chatgpt, gemini] difficulty: plug-and-play version: 1.0 source: original tested: no --- # Plain Email Rewriter ## What it does Rewrites any email into plain, direct, human language — strips corporate filler, apologies nobody asked for, and padding that buries the point. ## When to use it - Your draft sounds stiff or like a template - An AI wrote your first draft and it's 3x too long - You need to say something uncomfortable without hiding it in fluff ## The skill ``` Rewrite this email. Rules: - Say only what I actually want to say. Add nothing. - Cut every phrase that exists to sound polite rather than to communicate: "I hope this finds you well", "just circling back", "per my last email", "at your earliest convenience" - One idea per paragraph. Short sentences. - Keep it conversational, like I'd say it out loud to the person. - If my draft buries the main point, move it to the first line. - Same language and level of formality a normal person uses with a colleague they respect. Not a lawyer, not a robot. What I want to say: [YOUR ROUGH DRAFT OR BULLET POINTS] Who it's going to: [PERSON/ROLE] ``` ## Example output [TO FILL AFTER TESTING] ## Tweaks - Add "keep it under 80 words" for hard length control - Add "this is a follow-up, they haven't replied in a week" for context-aware tone - Works for Slack/LinkedIn messages too — just say the channel --- --- name: Voice Matcher category: content works_with: [claude, chatgpt, gemini] difficulty: needs-setup version: 1.0 source: adapted (user-submitted collection, rewritten) tested: no --- # Voice Matcher ## What it does Trains the AI on samples of your real writing so everything it drafts afterward sounds like you instead of a bot. Set it up once per project. ## When to use it - First message of any new chat/project where AI writes as you - Your AI drafts keep sounding generic no matter how you prompt - Handing content production to AI without losing your voice ## The skill ``` Below are [3-5] things I actually wrote myself — posts, emails, transcripts of me talking. Study them before writing anything. SAMPLE 1: [PASTE] SAMPLE 2: [PASTE] SAMPLE 3: [PASTE] From these, describe back to me in bullets: - My average sentence length and rhythm - Words and phrases I actually use - Words I would NEVER use - How I open and how I end things - My tone (blunt/warm/dry/etc.) Wait for me to confirm or correct that description. From then on, write everything in this voice. Rules that always apply: - Short sentences. No corporate words. No "delve", "unlock", "game-changer", "in today's fast-paced world". - If you're unsure whether I'd say something, leave it out. ``` ## Example output [TO FILL AFTER TESTING] ## Tweaks - Use voice-to-text transcripts as samples — that's your truest voice - Save the confirmed voice description and paste it into future chats instead of re-training - Redo it every few months; your voice drifts as you publish more --- --- name: Daily Three-Post Engine category: content works_with: [claude, chatgpt, gemini] difficulty: plug-and-play version: 1.0 source: adapted (user-submitted collection, rewritten) tested: no --- # Daily Three-Post Engine ## What it does Produces your daily social posts in one prompt — one that teaches, one that tells a story, one that starts an argument. Fifteen minutes a morning and the feed is fed. ## When to use it - Daily posting routine (Facebook, LinkedIn, X, Threads) - You post inconsistently because starting from blank is the blocker - Batching a week of content in one sitting ## The skill ``` Write 3 posts for today. Topic: [TOPIC] My audience: [WHO] | Platform: [WHERE] POST 1 — TEACH: One specific, usable thing. Not "5 tips" — ONE thing, done properly, with the how. POST 2 — STORY: First person, something that happened, told plainly. Real detail beats polish. End on the lesson without labeling it "the lesson". POST 3 — ARGUE: A question or take people genuinely disagree on within this topic. Pick a side lightly; leave room to fight in the comments. Rules for all three: - First line must work on its own — that's all most people see. - No hashtag spam (max 2, only if they earn it). No emojis unless I use them. No "Agree?" endings. - Each post stands alone. Someone seeing only one still gets value. ``` ## Example output [TO FILL AFTER TESTING] ## Tweaks - Stack with Voice Matcher first — this is where it pays off - Ask for 3 topics from your niche if even the topic is the blocker - "Make post 2 about [real event from my week]" beats invented stories every time --- --- name: Shorts Batch Week category: content works_with: [claude, chatgpt, gemini] difficulty: plug-and-play version: 1.0 source: adapted (viral "Shorts system" pack — 4 of 8 prompts merged into this one, 2 rejected for invented algorithm claims, 2 already covered) tested: no --- # Shorts Batch Week ## What it does Builds your complete short-video production system in one pass: a bank of ideas from content buckets, a batch filming session plan, one reusable editing template, and a two-week schedule with a built-in off day. The principle underneath: never do one video at a time — ideas, filming, and editing each happen in batches, so posting daily doesn't mean producing daily. ## When to use it - You post consistently for six days and vanish for three weeks (the burnout cycle) - Every video starts from zero: new idea, new setup, new editing decisions - Starting Shorts/Reels/TikTok and wanting a system before the chaos, not after ## The skill ``` Act as a video production planner for a one-person channel. Rules: build a system this one person can actually sustain — if my available time makes daily posting unrealistic, say so and plan the cadence my hours support; no claims about how "the algorithm" rewards anything — platform internals are unknowable from outside, so plan around consistency and viewer value, the things I control; judge performance only against my own numbers over time, never against invented benchmarks. My niche: [TOPIC + WHO IT'S FOR] Time I can really give this: [HOURS/WEEK, honestly] My setup: [PHONE/CAMERA, WHERE I FILM, EDITING APP] Talking to camera or voiceover: [WHICH] Build my batch system: 1. THE IDEA BANK: 20 video ideas from my niche using content buckets — myths people believe, mistakes I see, quick tips, stories from experience, reactions to common takes. Formatted as a reusable list I can refill by rerunning the buckets. 2. THE FILMING BATCH: how to shoot 5-7 videos in one session — grouping similar formats together, what to prep before pressing record (outfits, notes, framing), and how to mark takes so editing isn't archaeology. 3. THE EDITING TEMPLATE: one repeatable edit structure — where the hook text goes in the first second, caption size and placement that stays readable without covering the subject, my standard cuts and pacing — with a note on which elements stay identical for recognition and which vary so videos don't feel cloned. 4. THE TWO-WEEK SCHEDULE: batching, filming, editing, and posting mapped onto my real available hours, with one buffer day built in — a plan that survives one bad day without collapsing. ``` ## Example output [TO FILL AFTER TESTING] ## Tweaks - The idea bank empties faster than you think — refill it when 5 remain, not when it's empty mid-week - Steal your own best performer's structure for the editing template; recognition builds faster than novelty - After two weeks, run Video Retro on your real numbers — that's the honest version of "analytics," and it feeds the next batch - Pair with YouTube Hook Generator for the opening lines of each idea in the bank --- --- name: Winner Remixer category: content works_with: [claude, chatgpt, gemini] difficulty: plug-and-play version: 1.0 source: adapted (user-submitted collection, rewritten) tested: no --- # Winner Remixer ## What it does Takes your best-performing post and generates 5 fresh angles on the same core idea — because a post that already worked is a map, not a one-off. ## When to use it - Weekly, on whatever post did your best numbers - You found an idea that resonates and don't want to waste it - Reviving an old winner for a new platform ## The skill ``` This post did my best numbers this week: [PASTE POST] [OPTIONAL: numbers/comments it got] 1. First, tell me in one sentence WHY you think it worked — the core idea underneath the wording. 2. Then write 5 new posts on that same core idea, each a different angle: - Flip it (argue the opposite side of the same insight) - Zoom in (one detail from it, expanded) - Story version (same idea as something that happened) - Question version (make the audience answer it) - Contrarian version (attack the common advice around it) Rules: different words entirely — no recycled sentences from the original. Same voice. First line of each must stand alone. ``` ## Example output [TO FILL AFTER TESTING] ## Tweaks - Spread the 5 over two weeks; don't post them back to back - Also ask "which of these 5 would you bet on and why" — the reasoning teaches you your own audience - Works across platforms: a Facebook winner becomes 5 TikTok scripts