agentleFS
aya-talebaya-taleb/video-boost-skills/skills/viral-checklist/SKILL.md

viral-checklist

Skill1 starsChanged 22 days agoMITView on GitHub

Builds a checklist of what makes videos work right now for one topic on TikTok, Instagram Reels, YouTube Shorts and/or long YouTube videos, for a market and language (US and English by default). Runs the platform research skills, checks every source with source-check (12 months max, 6 with recent), and merges the findings into rules with a reason, a way to check, a weight and a source. Use when the user asks for a viral checklist, what makes X videos go viral, best practices for X on TikTok/Reels/Shorts/YouTube, or before video-score.

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 4 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    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
  1. Distinctive:

    Ask only for what is missing.

    line 12
  2. Distinctive:

    When findings conflict (for example two "best lengths"), write the rule as a range and say there is a conflict.

    line 32
  3. Distinctive:

    No rule without a dated source inside the window.

    line 42
  4. Distinctive:

    No numbers from memory.

    line 43

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Viral checklist

Turns fresh, checked research into a checklist that a person (or video-score) can test a video against.

Inputs

Ask only for what is missing. Defaults in brackets.

InputDefault
topicask
platformstiktok, reels, shorts (add youtube for videos over 3 minutes)
marketUS
languageen
window12 months (recent = 6 months)
format, if knownfor example narrated animation, talking head, skit

Steps

  1. Research each platform with its skill: tiktok-research, reels-research, shorts-research, youtube-research. They are independent, so run them at the same time when you can (for example as parallel subagents). Each writes notes-<platform>.json.
  2. Check sources with source-check on every notes file. Read source-report.md. If a platform has big gaps, tell the user before going on.
  3. Merge into rules. Group findings that say the same thing across platforms into one rule and list the platforms it applies to. Keep platform-only rules separate (for example Reels safe zones, YouTube end screens). For each rule fill the fields in shared/checklist-format.md:
    • level: must only for platform policy or high confidence evidence; avoid for mistakes; should for medium or high evidence that it helps; nice for low or mixed evidence.
    • check: how a person can tell, in one sentence. Use a measure (metric from video-score) whenever the rule can be measured.
    • evidence: the finding ids, so every rule traces back to dated sources.
    • When findings conflict (for example two "best lengths"), write the rule as a range and say there is a conflict.
  4. Add the topic's own risks, from the research (for example: book summaries must use your own words and credit the author; travel prices change; health topics give no personal advice).
  5. Write two files:
    • checklist.json (for video-score)
    • checklist.md (for people): grouped by stage (hook, story, length, pacing, visuals, text, audio, engagement, posting, policy), one line per rule with its level, and a Sources section with every link and its date. Set valid_until one month after today. Trends move fast; rebuild the checklist when it expires.
  6. Tell the user in a few lines: how many rules, the 5 that matter most, the gaps, and the date it expires.

Rules

  • No rule without a dated source inside the window.
  • No numbers from memory.
  • Market and language stay what the user set. If the only evidence is global, the rule says "global".
  • 15 to 40 rules is a good size. More than that and nobody uses it.
  • Plain words everywhere.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.