oliverhees/hermes-motion-pack

Step 1 of motion-pack. Runs a short, tap-friendly interview (max 2 rounds × 3 questions) and writes BRIEF.md for a motion graphics video — goal, audience, hook, format, length, must-show, sources, audio wishes. Use whenever a video is requested and no BRIEF.md exists yet, or the user says "Briefing", "interview me", "was brauchst du von mir".

Was ist hermes-motion-pack?

hermes-motion-pack is a Claude Code agent skill that step 1 of motion-pack. Runs a short, tap-friendly interview (max 2 rounds × 3 questions) and writes BRIEF.md for a motion graphics video — goal, audience, hook, format, length, must-show, sources, audio wishes. Use whenever a video is requested and no BRIEF.md exists yet, or the user says "Briefing", "interview me", "was brauchst du von mir".

Funktioniert mit~Claude Code~Codex CLI~Cursor

npx skills add https://github.com/oliverhees/hermes-motion-pack/tree/HEAD/skills/mp-brief

Dokumentation

mp-brief — interview, then brief

Goal: reach 98 % certainty with as few questions as possible, then write BRIEF.md.

Rules

Infer first, ask second. Read the request, the project folder, ~/motion-studio/STUDIO.json and any pasted material before asking anything.
Max 2 rounds. Round 1 = 3 questions. Round 2 only if something important is still open (max 3 more).
Offer options (A/B/C) so the user can answer with one word. Always add "du entscheidest" as an option.
Never ask what you can look up (website, docs, GitHub, screenshots).
Speak the user's language.

Round 1 (pick the 3 that matter most for this request)

Type / track — graphics: product spot · social reel · launch teaser · logo sting/loop · kinetic typography · data story · UI showcase — highlight: highlight reel from my footage · trailer/teaser · event recap/aftermovie; basis: my clips · AI clips · pure code · mix
Format + length — 16:9 1920×1080 · 9:16 1080×1920 · 1:1 1080×1080 · 4:5 1080×1350 · multi-format; 6 s / 15 s / 30 s / 60 s
Core message — one sentence the viewer should remember (ask for it, or propose 2 and let them pick)
Audience + platform — who sees it, where (YouTube, Reels/TikTok, LinkedIn, website hero, ads)
Audio — voice-over (which language, male/female/neutral) · music only · silent
Sources — website, docs, GitHub, screenshots, brand kit, existing assets (logo SVG!) — highlight track: footage folder/paths, music track (or mood + bpm), the "why" of the video (what should people feel after)

Round 2 (only if needed)

Hook (max 5 words) — propose 3, let them pick
Call to action
Hard constraints: brand colours, fonts, forbidden things, deadline
Reference videos they like (links)

Output: `BRIEF.md`

# BRIEF — <project>
Type: ...            Format: 1080x1920 @ 30fps, 15 s
Platform: ...        Audience: ...
Core message: "..."
Hook (≤5 words): "..."     CTA: "..."
Must show (in order of importance): 1. ... 2. ... 3. ...
Proof points (with source): - "..." (url)
Audio: voice de/female · music: energetic-minimal · SFX: yes
Brand: colours #..., fonts ..., logo path ...
Forbidden: ...
References: ...
Assumptions I made: - ...

Then present the brief in ≤ 8 lines and ask for OK. On OK: write STATE.json (phase: style), hand off to mp-style.

See references/interview-bank.md for ready question sets per video type.

Files in this skill

references/interview-bank.md