create-shortform
Skill14 starsChanged 3 months agoMITView on GitHub
Research, brief, script, and evaluate short-form video. Use for TikTok, Instagram Reels, YouTube Shorts, founder demos, UGC-style concepts, hooks, storyboards, recuts, production specifications, or learning from watch and qualified-response data.
In this repository22 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- brief-creativeSkill
- brief-product-uiSkill
- conquistadorSkill
- create-brandSkill
- create-paid-campaignSkill
- debate-agentsSkill
- improve-conversionSkill
- knowledge-reviewSkill
- map-user-flowSkill
- measure-growthSkill
- model-growth-funnelSkill
- optimize-searchSkill
- plan-campaignSkill
- polish-vietnameseSkill
- research-channelSkill
- research-positioningSkill
- write-copySkill
- write-longformSkill
- write-outreachSkill
- write-socialSkill
The file
We split this skill into the rules it gives an agent
2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 18
When current platform or trend evidence matters, inspect a dated sample large enough to distinguish a pattern from an anecdote.
- Distinctive: line 26
Do not claim a trend from memory, manufacture VoC, or treat another account's benchmark as the user's baseline.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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
Create a short-form video package
Build a production-ready short-form argument, not a generic script pasted into every vertical feed.
Establish evidence and hypothesis
Define audience, costly moment, one communication job, product mechanism, proof, desired action, and the behavior the video should test.
When current platform or trend evidence matters, inspect a dated sample large enough to distinguish a pattern from an anecdote. Separate:
- durable attention/story mechanics;
- current platform formats and eligibility rules;
- ephemeral audio, edit, or trend signals;
- owned performance evidence.
Do not claim a trend from memory, manufacture VoC, or treat another account's benchmark as the user's baseline.
Script and storyboard the hero
Make the first two seconds carry a specific tension, result, or product behavior. Every following beat must develop the promise, demonstrate the mechanism, resolve an objection, or deliver the payoff.
Provide timed voice/dialogue, on-screen copy, visual action, audio intent, captions, CTA, and the source product/brand evidence. Use real UI or label representative material. Never invent product behavior, users, metrics, or screen states.
Create true platform recuts
Adapt the edit, hook, pacing, title/caption, payoff timing, and safe zones for each selected platform. A crop or renamed export is not a recut. Verify current platform specifications before production.
Specify capture, aspect ratio, resolution, typography, audio, privacy/redaction, source files, and fallback when rendering or product capture is unavailable.
Evaluate and learn
Evaluate one platform/version at a time using the source hypothesis. Record sample/window, hook hold, retention/completion, qualified response, baseline comparability, and confounders. Do not let views override a failed business or audience signal.
Finish with keep/drop/change, one next hypothesis, and what the next brief should do differently.
Deliver
Return the evidence boundary, audience/hypothesis, hero script, storyboard, platform recuts, production package, measurement plan, and external-action boundary. Posting and paid amplification remain human-owned.
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.

