Pull Request Explainer Video

Plugin installs: Before setup or freshness commands, follow plugin execution rules when this skill is inside a HyperFrames plugin. Standalone installs keep the update instructions below.

First, keep this skill fresh — confirm with the user before running: npx hyperframes skills update pr-to-video. A fast no-op when everything is current; otherwise it refreshes this skill plus the core domain skills it depends on before you rely on them.

media-use: Before sourcing audio/images/logos, call /media-use to resolve BGM/SFX/images from the HeyGen catalog and brand logos from their official sources. Run --adopt first to register existing assets. See /media-use skill. When the host app provides its own music or sound-effect tools, use those for music and sound effects.

PR to HyperFrames

Use this skill to ingest a GitHub pull request, understand the change, plan a code-change explainer, and build it frame by frame in HyperFrames. The input is a code change (read via gh), not a website — there is no capture step and no real assets beyond the contributors' avatars.

The front door is /hyperframes. You are the orchestrator. Run each step, verify its gate, and only then continue. This skill is for a GitHub pull request (a code change). Any other intent, a bare "make a video", or any uncertainty → read /hyperframes first — the intent layer owns every route decision, and a fresh creation arriving here without BRIEF.md goes through it anyway (Setup's opening rule).

You are the orchestrator. Work in the resolved external PROJECT_DIR, never in the caller repository by default. Run steps in order and pass each gate before continuing. User-gated steps are Step 0, Step 3, and Step 6. Read ../hyperframes/references/brief-contract.md before Step 0 — it defines the gate types and how BRIEF.md's flow/storyboard derive the mode that governs the Step 3/4/6 gates. Do every step yourself except Step 5, where you dispatch a bounded pool of frame workers. Do not put design or motion rules here; those live in the frame-worker sub-agent, this skill's local ../hyperframes-animation/rules/ + ../hyperframes-animation/blueprints/, and hyperframes-creative.

Workflow: Step 0 setup → hyperframes.json; Step 1 ingest → capture/extracted/ + assets/<login>.png; Step 2 design system → frame.md; Step 3 storyboard/script → STORYBOARD.md and SCRIPT.md; Step 3.1 audio → audio_meta.json; Step 4 visual design → enriched STORYBOARD.md; Step 5 frames → compositions/frames/NN-*.html and index.html; Step 6 final render → renders/video.mp4.

Step 0: Setup

Goal: Enter with a confirmed brief — including the PR reference (a full URL, an <owner>/<repo>#<N> ref, or "this PR" in a checked-out repo) — create the HyperFrames project, and make the brief durable. The style is always code-editorial (fixed at Step 2, never asked).

The brief is confirmed by the intent layer, not by questions asked here. Opening rule, in order: (1) BRIEF.md exists → read it and ask nothing — the brief is settled, and its flow/storyboard derive the mode (brief contract § 1). (2) No BRIEF.md but the project exists (hyperframes.json / STORYBOARD.md on disk) → resume from the storyboard's frontmatter and the recorded preferences; never re-interrogate a half-built project. (3) Neither — a fresh creation request that arrived here directly → read /hyperframes and run its intent layer (references/intent-interview.md): it checks recipes and remembered defaults, and conducts this route's questions — including the PR-size → length doctrine, which lives whole in ../hyperframes/references/routes/pr-to-video.md — then hands back the locked brief. Edit requests skip all of this — go do the edit.

Resolve the project directory before doing any other work. Preserve a user-supplied project directory; otherwise use the durable external cache location printed by the resolver. Never create videos/ in the caller repository:

PR="<url | owner/repo#N>"
if [ -n "${EXPLICIT_PROJECT_DIR:-}" ]; then
  PROJECT_DIR="$(node <SKILL_DIR>/scripts/project-dir.mjs --pr "$PR" --project-dir "$EXPLICIT_PROJECT_DIR")"
else
  PROJECT_DIR="$(node <SKILL_DIR>/scripts/project-dir.mjs --pr "$PR")"
fi
echo "PR-to-video project: $PROJECT_DIR"
node <SKILL_DIR>/scripts/preflight.mjs

The capability preflight runs before fetch, story work, audio, or frame dispatch. If the installed CLI cannot run the validation command required by this skill, stop with its upgrade instruction rather than spending the run's context first.

Initialize only if $PROJECT_DIR/hyperframes.json is missing. Its basename comes from the PR, such as acme-sdk-pr-1842; never use the workspace name or a timestamp.

npx hyperframes init "$PROJECT_DIR" --non-interactive --example=blank --skill=pr-to-video — init checks the installed skills against the latest on GitHub and updates the global set if any are out of date.

Every relative-path command below runs with $PROJECT_DIR as its working directory. Examples without an explicit subshell mean (cd "$PROJECT_DIR" && …); never change the caller repository's working tree.

Write BRIEF.md immediately after init (never before — init refuses a non-empty directory): the intent layer's locked brief, shape per ../hyperframes/references/brief-format.md. Resolve <MEDIA_DIR> as the installed /media-use skill directory. Then record each preference-backed answer with node <MEDIA_DIR>/scripts/prefs.mjs record --hyperframes . (brief-format.md names the subset). If the intent layer adopted a recipe, run node <MEDIA_DIR>/scripts/recipe.mjs use --hyperframes . --name <name>; it copies its frame.md into the project (Step 2 is then skipped) and returns the skeletons Step 3 drafts from. A recipe fills answers, not approvals; the review gates still run.

Show sign-in status before proceeding past Setup — run npx hyperframes auth status and relay its output verbatim. It reports whether voice/BGM will use HeyGen or local engines and, when signed out, how to sign in. Apply one branch:

• Collaborative: wait for the user to sign in or explicitly choose offline / go.
• Autonomous: state the status and continue through the available local engines.

Do not silently omit a required capability when no offline provider exists; surface the blocker. Do not fold this decision into another question or write keys into a per-repo .env. Auth ownership and offline fallbacks: /media-use references/setup-providers.md § Providers.

Gate: hyperframes.json and BRIEF.md exist; the PR ref is captured in the brief; the preference-backed answers were recorded (brief contract § 2); sign-in status was shown (signed in, or continuing offline).

Step 1: Ingest the PR (no capture)

Goal: Fetch the PR's facts and fold them into the project as the source of information. There is no website capture. fetch-pr.mjs runs gh deterministically — completing the files list via paginated gh api so a large PR doesn't truncate at ~100 files, and writing only capture/pr.json + capture/diff.patch (no scratch dir). For MERGED PRs it also resolves a best-effort shipped_version (+ version_source) into pr.json, so the end card can cite a real version instead of inventing one. Then ingest.mjs folds that into the synthetic capture package offline.

PR="<url | owner/repo#N | N>"

# Fetch the PR deterministically: runs gh, completes the files list via paginated
# gh api (so a big PR doesn't truncate at ~100 files), writes only capture/pr.json +
# capture/diff.patch — no scratch dir. gh auth / not-found / private errors exit 1 here.
(cd "$PROJECT_DIR" && node <SKILL_DIR>/scripts/fetch-pr.mjs --pr "$PR" --out-dir ./capture)

# Offline transform → capture/extracted/{tokens.json (colors:[] → code-editorial palette),
# visible-text.txt (the brief), people.json (contributors, bot-filtered, name+login,
# avatarFile=assets/<login>.png)}.
(cd "$PROJECT_DIR" && node <SKILL_DIR>/scripts/ingest.mjs \
  --pr-json ./capture/pr.json --diff ./capture/diff.patch --out-dir ./capture/extracted)

# The people front's one network step — download each contributor's GitHub avatar to
# assets/<login>.png for the credits close. Best-effort; always exits 0.
(cd "$PROJECT_DIR" && node <SKILL_DIR>/scripts/fetch-people-avatars.mjs \
  --people ./capture/extracted/people.json)

If fetch-pr.mjs exits 1 (gh auth / not found / private), report its stderr and stop — do not fabricate PR contents. If ingest.mjs exits 1, read its stderr (usually a malformed pr.json), fix, and rerun (deterministic). fetch-people-avatars.mjs always exits 0; missing avatars just mean no credits close to author.

people.json carries a name for whichever contributors gh already named (the PR author, commit authors, mergedBy) — null for the rest (reviewers/commenters/assignees, which gh pr view only ever gives a bare login). Before writing the credits close in Step 3, resolve any null name yourself for the 1-6 people who'll actually appear on that frame: gh api users/<login> --jq .name (you already have gh — no need to script this). If GitHub has no public name for that user either, fall back to the login on-screen and drop that person from the spoken line (see story-design.md's credits section — the voiceover must still say names, never raw handles).

Gate: capture/pr.json, capture/diff.patch, capture/extracted/tokens.json, capture/extracted/visible-text.txt, and capture/extracted/people.json exist; you can state the PR's change in one clear sentence. assets/<login>.png is best-effort — its absence is not a failure.

Step 2: Design System

Goal: Adopt the code-editorial frame preset; a script turns it into this video's frame.md + caption skin.

The style is fixed — code-editorial (warm editorial; a navy code surface built for diffs). Run:

node <SKILL_DIR>/scripts/build-frame.mjs --preset code-editorial --hyperframes .

The script copies the code-editorial preset's FRAME.md → frame.md, remixes it onto any brand tokens in capture/extracted/tokens.json (a PR has none → colors:[]/fonts:[] keeps code-editorial's own palette, a complete design), copies the preset's caption skin to .hyperframes/caption-skin.html, and self-validates (exits 1 on a broken mapping). Proceed as soon as it exits 0 — no hand-editing.

Gate: build-frame.mjs exited 0 — frame.md exists from the code-editorial preset, and .hyperframes/caption-skin.html exists as the caption skin source.

Step 3: Storyboard and Script

Goal: Turn the PR into an approved frame-by-frame explanation plan.

Read ../hyperframes-creative/references/story-spine.md (hook language, value-before-evidence, storyboard-as-proposal, source-traceable visuals), references/story-design.md, ../hyperframes-animation/blueprints-index.md, ../hyperframes/references/storyboard-format.md, and ../hyperframes/references/script-format.md. Use them to write STORYBOARD.md and, when narration is needed, SCRIPT.md. Set the frontmatter duration: from the brief's length — a rough expectation; assembly reports where the cut lands against it.

Use story-design.md for the PR archetype (changelog / feature-reveal / fix-explainer / refactor-walkthrough), the PR-native frame types, hook, persuasion, beats, the per-frame word budget, and the credits close. The sequence comes from narrative design, not the diff's file order — explain the change, don't read the diff aloud. As a soft guide, consult the role→blueprint menu in ../hyperframes-animation/blueprints-index.md: for each beat, write the voiceover in the shape its candidate blueprint implies and tag that candidate blueprint: id when one fits (story truth still decides which beats exist — never force a beat to fit a shape). Feature 2–4 real diff hunks (from capture/diff.patch), each a small legible snippet; name the code-* block each wants in the frame's scene. Frames carry no asset_candidates except the credits close (1–6 assets/<login>.png avatars). Use the exact required fields from the storyboard and script references.

After drafting, run the review loop's plan pass — ../hyperframes/references/review-loop.md § 1: present the plan as a proposal, and ask the two questions — approve or change, and sketches first (recommended) or skip. Feedback arrives as a chat reply; loop until approved. This is a checkpoint gate (brief contract § 1): in autonomous mode there is nothing to ask — post the same summary as a heads-up and proceed; sketches collapse into the build, and the final preview with its one question comes at Step 6.

Gate: STORYBOARD.md exists, every frame has the required narrative fields, SCRIPT.md exists when narration is needed, and the user approved the plan (autonomous: the summary was posted as a heads-up).

Step 3.1: Audio

Goal: Generate narration, word timings, music, and audio metadata from the approved script.

Start audio after Step 3 approval. Run it in the background, then continue to Step 4.

Choose the narration voice from the user's ask before invoking. If the request named a voice, gender, or tone, pick a matching voice id and pass it with --voice <id>. The pipeline default is otherwise Marcia (female) on HeyGen / am_michael on Kokoro — so a request like "a male voice" is silently ignored unless you pass the flag. Voice ids are provider-specific; resolve against whichever provider Step 0's sign-in status selected: HeyGen (signed in) via node <MEDIA_DIR>/audio/scripts/heygen-tts.mjs --list (or GET /v3/voices?engine=starfish); Kokoro (offline) via the voice table in <MEDIA_DIR>/audio/references/tts.md (prefixes am_/bm_ male, af_/bf_ female). When the user expressed no preference, fall back to the remembered voice (brief contract § 2) before the pipeline default, and say which one you used; omit --voice only when neither names one. When the user explicitly picked a voice this run, record it (prefs.mjs record --key voice).

node <SKILL_DIR>/scripts/audio.mjs --script ./SCRIPT.md --storyboard ./STORYBOARD.md --hyperframes . --out ./audio_meta.json --voice <voice-id> &

The audio script handles narration, word timings, BGM lookup from HeyGen's music library, and timing metadata. BGM mood comes from the storyboard's music: field. This uses the HeyGen Audio API for retrieval, not generation, and the same ~/.heygen credential as TTS. For provider details, read ../media-use/audio/references/tts.md. When the host app's own tools make the music or a sound effect, put the file under assets/ and list it in the audio_meta.json file marked "source": "host": music as bgm { "path": "assets/…", "source": "host" } with music: none in the storyboard; a sound as an sfx entry { "frame": <n>, "file": "assets/…", "offset_s": 0, "duration_s": <its length>, "source": "host" }. Every audio pass keeps these, and a frame with a host sound gets no looked-up cue.

If there is no narration and no SCRIPT.md, skip voice generation. BGM may still run if the storyboard has a music mood.

The canonical fully-silent marker (shared across the workflows that reuse this audio model): music: none in the STORYBOARD.md top YAML block, no SCRIPT.md, and no sfx: cues or host audio. That combination marks the project silent — no narration, no BGM, no SFX. audio.mjs recognizes it and generates nothing (it removes any stale audio_meta.json; an absent audio_meta.json is what assemble treats as silent), so this step is a clean skip. Without narration but with sfx: cues or host audio the film is not silent: audio.mjs writes the audio_meta.json file for fetch-sfx to fill. music: none with narration keeps TTS and turns only BGM off. Use exactly this spelling — don't improvise other markers.

Gate: audio job has started, or the project is marked silent (music: none, no SCRIPT.md, no sfx: cues or host audio).