Papercut Storyboard

Before you start: read this whole install prompt, tell me in a few lines what this skill does and what installing it involves, and wait for my go-ahead before you do anything else. Prerequisites: none for the install itself: no account, no API key, no paid service, and nothing else to install first. This listing installs one text file only — SKILL.md, no code, no scripts. For using the skill afterwards, read this first, because this skill is NOT self-contained: SKILL.md says "Minimal is enough", but its full workflow leans on things that live outside its folder and are NOT installed by this install (all disclosed in point 2): the shared papercut library (its catalog, its kit components paper/camera.tsx, paper/transitions.tsx and paper/kit.tsx, its backdrop sheet and its audio catalog), two reference files from the sibling skill papercut-explainer (references/style-guide.md and references/motion-principles.md), and the neighbouring pipeline stages papercut-hook and papercut-script before it and papercut-build after it. In practice you need the rest of the papercut-explainer repository set up (the repository's own setup route writes the library path into ~/.config/papercut-explainer/config.env) to use it as written. I also provide the project inputs myself, in my own episode project: narration/script.json and papercut/STORYBOARD.md already carrying the HOOK block and the staging table. One use-time step reaches the network, and only that one: the once-per-reel proof beat screenshots a real source page through an external library tool (library/tools/capture.mjs, not part of this folder), which visits the source page's URL in a headless browser — disclosed in point 2 before it ever happens. Install "papercut-storyboard" for me. It storyboards a papercut script — a short vertical explainer that looks like layered construction paper — before any code is written. It puts the global rules at the top of STORYBOARD.md (1080x1920, 30 fps, banner at 14-21% height, caption band 68-80% shifting per scene to stay off faces, bottom 20% clear, light from the top-left, saturation only on characters, captions and signals, palette/fonts/canvas taken from the sibling style guide), writes the Camera / Cut table (one row per scene, every row naming a component from the kit's paper/camera.tsx — OrbitCamera + Depth, SnapZoom, Crane, CrabTrack, DollyZoom, RackFocus and the rest — and the CutSeries kind from paper/transitions.tsx for the cut into that scene — hard, whip, zoomThrough, pageTurn, tearAway, tapePull and the rest — with a plain CameraPush on at most a third of the scenes, each row's camera copied into its scene block header and each cut copied into episode.json as a "cuts" entry), then one scene block per scene (1.5-5 s each, cutting on sentence boundaries: banner takeaway, location/background, camera, cut in, cast, caption band, a frame-level event table with one event per stressed word every 0.4-0.8 s, ambient motion always on, boil on banner and labels, and the anticipation plus the one exaggerated beat named in each block), the SFX plan (sounds named from the library audio catalog, air sounds at most one every 2.5 s, a bed preset and cutSfx matching the cut kinds), exactly one proof beat per reel (a real screenshot of the source page, taped onto a paper card with a red marker circle around the key line — everything else stays a rough papercut rebuild of the product UI on a paper monitor or laptop) and one callback to an earlier set before the CTA, and it checks the closing gate (every scene has a camera and a cut-in, every stressed word has an event, at most one page turn or tear per reel, air sounds within budget). Repository: https://github.com/0xS4N4TG/papercut-explainer (the skill is the whole skills/papercut-storyboard/ folder at the pinned commit: SKILL.md only, 1 file) Pinned version: install the code at commit ff8223abc8136cc09e8d7e9b62ecb5d080c389e1 exactly. Do not substitute a newer version. If this commit is unreachable, stop and tell me. 1. Clone https://github.com/0xS4N4TG/papercut-explainer into a temporary folder, check out commit ff8223abc8136cc09e8d7e9b62ecb5d080c389e1, and copy the skills/papercut-storyboard/ folder from that checkout. Then compare every file in the folder with the fingerprints recorded for this listing on 2026-10-19: [FINGERPRINTS] - SKILL.md: SHA-256 6b69d53f7ae5cbee47008c6e9f2fafe946de5912bf3ac2d72daab709cb8329e5 If a file does not match, it changed since the listing was prepared: do not install it yet, tell me what you found and let me decide. 2. Read SKILL.md in full, and tell me in plain words: what the skill does when used, which services it connects to, whether it reads or sends any personal information or keys, and anything that seems unusual. Stop on any red flag and tell me before going further. Confirm these points from the file: - The skill folder holds 1 file only, plain text: SKILL.md (6,933 bytes). There is no code, no script, no reference file, no import and no URL anywhere inside the folder. Everything SKILL.md points at lives outside the folder (next point) — this skill is a planning brief that names other files, components and stages; it does not bundle them. - Dependencies outside the folder, disclosed here — none of these is installed by this install, and do not fetch, install or run any of them during this install: (a) the shared papercut library, resolved at use time from the environment variable PAPERCUT_LIB, else from the PAPERCUT_LIB= line in ~/.config/papercut-explainer/config.env (a config file this skill reads at use time), else by asking me once — from that library SKILL.md reads library/papercut/CATALOG.md (the "paper/camera.tsx" and "paper/transitions.tsx" sections), opens library/papercut/public/bg/backdrops/_sheet.jpg to choose sets, names its camera and transition components from the kit files library/papercut/kit/paper/camera.tsx, transitions.tsx and kit.tsx, and names its sounds from library/audio/CATALOG.md; if the catalog file is missing or stale, SKILL.md says to run python3 $PAPERCUT_LIB/tools/catalog.py (the library's own tool, outside this folder, local-only file scanning per its imports). (b) Two reference files from the sibling skill folder skills/papercut-explainer/, reached as ../papercut-explainer/references/style-guide.md (section 5 for palette by role, section 6 for fonts, section 4 for canvas and zones, section 12 for a feature-unlock metaphor) and ../papercut-explainer/references/motion-principles.md (the 12 animation principles mapped to kit components). (c) The neighbouring stages, none of them part of this listing: it needs narration/script.json and a papercut/STORYBOARD.md whose HOOK block and staging table come from the sibling skills papercut-hook and papercut-script (SKILL.md says to run those two if the inputs are missing), it applies papercut-hook's rule to the cut out of the hook in S01 and S02, and its hand-off is the sibling skill papercut-build, which scaffolds the project and fills episode.json from the Camera / Cut table. (d) My own episode-project files, which are inputs and outputs, not repository files: narration/script.json, papercut/STORYBOARD.md, episode.json, tools/cues.json and tools/cue_sheet.d/<Scene>.json (the scene agents write the real SFX rows there, using the library's cues.py counts). - Nothing is sent anywhere during installation itself; the only download at install time is the skill folder from GitHub at the pinned commit. No script is run during installation, no key is asked for, and there is no telemetry in the folder — SKILL.md contains no URL at all. - What leaves this machine at use time, disclosed here before it ever happens: the storyboard writing itself sends nothing — it is local work between you and me on my project files (reading narration/script.json and STORYBOARD.md, writing the global rules, the Camera / Cut table and the scene blocks into STORYBOARD.md, and the cut entries shaped for episode.json). The one exception, named in SKILL.md's scene-block rules: the proof beat, once per reel, uses a real screenshot of the source page, captured via library/tools/capture.mjs — a tool from the shared library, outside this folder and not installed here — which launches a headless Chrome/Chromium through puppeteer-core and visits the source page's URL (the page my script is sourced from), downloading that page to screenshot it locally as a PNG for the later build. That visit, to that page's host, happens only at use time, only for the proof beat, and only when the library is present and that later capture step is actually run; this skill's own text performs no network call, names no API, no account and no key. If I have no library, or skip the proof capture, nothing leaves this machine when this skill is used. Later stages (papercut-build, papercut-voice, rendering) are separate skills with their own behaviour, not this one's. - The honesty rules from SKILL.md: product UI is always the rough papercut rebuild on a paper monitor or laptop with fake interactions — the single proof screenshot is the only real capture, taped onto a paper card, never a redraw passed off as real; a plain CameraPush is allowed on at most a third of the scenes; at most one page turn or tear per reel; transitions run 6-12 frames, each with its SFX. - Do not install the library, the sibling skills or anything else during this install, do not run catalog.py, capture.mjs or cues.py, do not create an episode project, and do not write any storyboard: using the skill is my decision, later, when I actually ask for one. 3. When I say "yes, install", install it as a skill: copy the whole papercut-storyboard folder into the agent's skills directory, in a folder named "papercut-storyboard", keeping its 1 file (SKILL.md at its root). A manual clone followed by copying the skills/papercut-storyboard folder into ~/.claude/skills/ is the route described here; any other route is fine only as long as every installed file matches its fingerprint. 4. Verify with no network calls: frontmatter valid (name: papercut-storyboard), the 1 file in place with matching fingerprint. Do not run any script to verify — there is none in this folder. 5. Report what was installed, where, and what I still need to do myself (for example: set up the rest of the papercut-explainer repository separately if I want the skill to work as written — the shared library with its catalog, kit, backdrop sheet and audio catalog, and the sibling skills papercut-explainer (for its two reference files), papercut-hook and papercut-script (for the HOOK block, staging table and narration/script.json), with PAPERCUT_LIB resolvable from my environment or from ~/.config/papercut-explainer/config.env; provide narration/script.json and papercut/STORYBOARD.md in my episode project; expect the output as the global rules, the Camera / Cut table and one scene block per scene in STORYBOARD.md, ready for papercut-build; using the skill is a separate step that only starts when I ask for it, and the only use-time network route disclosed above — the proof-beat capture of my source page through the library's capture.mjs — happens only then, if that step runs). GitHub is optional: if I have a GitHub account or the gh CLI, you may use it; otherwise public access is fine. Never require it unless it's in the prerequisites above. Rules: don't touch anything outside the temp folder and the install target. If anything looks off, stop and ask me.