clip-to-canon-finder

Identify short-form moments most likely to turn into fandom artifacts, edits, or memes from clips, transcripts, and reaction data, then output a shortlist with reuse angles. Use when Codex needs to fi

SKILL.md

GitHub

Clip to Canon Finder

Use this skill when you need to score short-form moments by meme potential, editability, and canon value.

Default product shape: a compact macOS MenuBarExtra utility with a source-aware popover for clip intake, moment scoring, and shortlist export and an optional settings/history window.

Quick Start

  1. Confirm the subject, target audience, and source window.
  2. Gather only public or user-authorized material into a working set.
  3. Produce a clip-shortlist with the default sections below.
  4. If the user wants a richer product concept, keep it menu bar first and native to SwiftUI.

Accepted Inputs

  • video links, rough timestamps, or short-form exports from Reels, Shorts, or TikTok
  • transcripts, subtitle files, or manual clip notes
  • comment samples that show which line or reaction fans keep repeating
  • editor notes about current formats, cut length, and reuse constraints

Output Artifact

Primary artifact: clip-shortlist

Default sections:

  • Shortlisted moments with timestamps or identifiable cut points
  • Why each moment sticks: line, look, reaction, or dynamic
  • Fandom reuse angle for edits, memes, callbacks, or compilation posts
  • Cut guidance for short-form packaging and thumbnail framing
  • Moments to skip because the signal is weak, mean-spirited, or misleading

Workflow

Gather Sources

  • Normalize clip references and match them to transcript segments when available.
  • Mark where the audience is quoting, clipping, or repeatedly reacting to the same beat.
  • Separate the raw moment from platform-specific editing noise.

Distill Signal

  • Score moments by surprise, clarity, quotability, and replay value.
  • Differentiate wholesome fan canon from rage-bait or context collapse.
  • Identify which moments can support sequels, remixes, or callback edits.

Build The Artifact

  • Export a shortlist with timestamp, hook, signal strength, and reuse angle.
  • Suggest one clean cut pattern for each winning moment.
  • Flag any clip where quote-mining or decontextualization would mislead the audience.

Mac Product Shape

  • Prefer MenuBarExtra with these primary actions:
    • Add Clips
    • Score Moments
    • Review Shortlist
    • Export Angles
  • Keep the popover between 320 pt and 380 pt wide with grouped sections, clear source status, and a strong export action.
  • Use a lightweight settings or history window only for source rules, saved exports, or weighting controls.

Safety Boundaries

  • Work only from public or user-authorized material.
  • Do not assist with stalking, covert profiling, rumor laundering, harassment, or minors-focused manipulation.
  • Do not recommend deceptive quote-mining, harassment bait, or edits that distort the original context in a harmful way.

Example Prompts

  • Use $clip-to-canon-finder to score these livestream moments and tell me which ones can become repeatable edit bait without feeling forced.
  • Use $clip-to-canon-finder to analyze these founder short-form videos and shortlist the moments worth turning into recurring canon callbacks.

Resources

  • references/product-spec.md: product surface, SwiftUI layout, artifact template, icon direction, and first-run behavior.