Claude Skill

course-module

Online course lesson editor for Claude: turns a teacher's recording (talking head, slides or screen) into a clean 3-10 min lesson with one learning objective, numbered steps, callouts, quiz cards, a recap, accessible captions (WCAG) and a transcript, consistent across every modul

LLM Mart · 0 points · 0 views 0 listing impressions 0 install-command copies
Download viprasol-tech-claude-video-editing-skills-plugins_video-studio_skills_course-module-0293497.zip · 3 KB

Install

skills CLI npx skills add https://github.com/Viprasol-Tech/claude-video-editing-skills/tree/main/plugins/video-studio/skills/course-module
Claude Code claude plugin marketplace add https://llmmart.ai/marketplace.json && claude plugin install viprasol-tech-claude-video-editing-skills@llmmart
Git git clone https://github.com/Viprasol-Tech/claude-video-editing-skills.git

The skills CLI installs just this skill, for any of its supported agents. Claude Code installs the whole viprasol-tech/claude-video-editing-skills collection as a plugin from our marketplace. Git is the plain clone.

Skill manifest

Course Module

One lesson, one objective, nothing the learner has to rewind to understand.

  • Best for: online courses, corporate training, onboarding lessons, university lectures, coaching programmes
  • Inputs: lesson recording (face, slides, screen or all three), slide deck export or screen capture, course name
  • Default skin: whiteboard-host · alternatives tutorial-steps, highlighter-study, chapter-pro
  • Formats: 16:9 lesson; optional 9:16 promo clip per module
  • Durations: 3-10 min per lesson (split longer recordings)
  • Genre preset: --genre course-module

Before you start

  • Needs video-studio-core in the same skills folder. CORE = that folder, SKILLS = the skills folder.
  • First run: python CORE/scripts/check_env.py.
  • Ask for: the learning objective of this lesson (one sentence), terms and names to spell (--prompt), the course template (intro card, colours) so every module looks the same, and whether an LMS needs SCORM.
  • SCORM packaging is not done by the engine: deliver MP4 + VTT + transcript and let the LMS wrap them.
  • Accessibility is part of the job: captions on every lesson, transcript text, nothing conveyed by colour alone.

How a pro cuts this

Structure (% of lesson)

beat % what goes here
objective 8 "By the end of this lesson you can..." on screen and said
why it matters 10 the problem this solves for the learner
teach 55 concept in 2-4 numbered steps, one idea per step
example / practice 17 worked example, then a quiz card or exercise prompt
recap + next lesson 10 3 bullet recap, what comes next

Pacing. ASL 4-10 s; visual change every 8 s (callout, step card, slide change). Slower than YouTube: learners need time to read. Hold slides with text >= 1.5 s per line of text.

Cut rules.

  • Keep the last good take (--retakes); remove fillers and long pauses but keep the teacher's rhythm.
  • Jump cuts are acceptable in lessons; hide the obvious ones with a slide or a punch-in.
  • Cut repetition and tangents; move asides to a "notes" document, not the video.
  • Split recordings at topic shifts into separate lessons (story.py outline shows topic shifts).

Transition grammar. Cuts; a dissolve only at a scene change (talking head to slides). Same intro and outro card on every module.

Music and sound. No music under teaching (bed at -32 dB only on the intro/outro card). Clear voice: "eq": "clear", de-ess, -16 LUFS for LMS/web. No SFX except a soft tick on a quiz reveal.

Colour. No look (accurate). Fix white balance and exposure only (auto correction).

Captions. Accessible subtitles, sentence case, <= 15 cps, 2 lines max, high contrast ("preset": "clean"); always also deliver .vtt/.srt sidecars and a transcript.

What the client judges. Clarity, consistency across modules, accessibility.

Common mistakes. No stated objective; 25-minute lessons; reading slides aloud; music under explanation; captions too fast; every module with a different look; missing transcript.

Make the edit

python CORE/scripts/run.py lesson03.mp4 --style SKILLS/whiteboard-host --genre course-module --format 16:9 --smart-cuts --retakes --prompt "Kubernetes, kubectl, YAML" --title "Lesson 3: Deployments" --work m03 --stop-after plan
python CORE/scripts/story.py outline --work m03 --genre course-module    # topic shifts -> split long recordings
python CORE/scripts/render.py m03/plan.json --frames 2,30,90,180,280
python CORE/scripts/render.py m03/plan.json
python CORE/scripts/finish.py m03/plan.json --platform podcast

--platform podcast gives -16 LUFS (LMS/web target). Slides or screen as a separate file: add --clip slides.mp4 and place it with clip elements or full-frame cutaways. Split a long recording into lessons with --cut ranges or story.py reorder --work m03 --order "s1-s24" per lesson, then run.py - --timeline ... per lesson.

Refine the plan

  • Objective card: {"type": "titlecard", "kicker": "Lesson 3 of 8", "title": "Deployments", "subtitle": "Roll out a new version with zero downtime", "start": 0, "end": 4}.
  • Steps: {"type": "steps", "steps": [{"label": "Write the YAML", "at": 31}, {"label": "Apply", "at": 74}, {"label": "Verify", "at": 120}]} or a checklist with items[].at when the teacher lists things.
  • Callouts and highlights: {"type": "callout", "point": [880, 420], "x": 1450, "y": 220, "title": "replicas: 3"}, {"type": "text", "text": "rolling update", "fx": {"name": "highlight", "target": "rolling"}}.
  • Quiz card: {"type": "poll", "question": "Which command applies a manifest?", "options": [{"label": "kubectl apply", "pct": 100}, {"label": "kubectl run", "pct": 0}], "revealAt": 3} (pause cue: hold 4-5 s), or a card with title and items.
  • Recap: card with 3 items timed to the teacher's recap words.
  • Captions: {"type": "captions", "preset": "clean", "cpsMax": 15}; fix terms in words.
  • Audio: "audio": {"platform": "podcast", "voice": {"denoise": "light", "eq": "clear", "deess": true}}; no music (or only on the intro card with "fadeOut": 1). Remove sfx the skin added.
  • Colour: "color": {"auto": true} only; no look.
  • Keep a course_template.json of the intro card, colours (theme.vars) and caption settings; paste it into every module.
  • python CORE/scripts/editor.py review m03/plan.json --genre course-module.

Deliver

python CORE/scripts/deliver.py m03/plan.json --to youtube,web-h265 --out-dir m03/deliver --burn no --cps 15 --chapters "0:00 Objective; 0:40 Why; 1:30 Steps; 5:10 Practice; 6:40 Recap"
python CORE/scripts/qc.py m03/deliver/final_youtube.mp4 --plan m03/plan.json --platform youtube --strict

Hand over per lesson: MP4, captions.vtt + .srt, a transcript (the caption text from post.md or the words), chapters. Name files course_m03_deployments_v002.mp4. SCORM: give the LMS these files; the engine makes no package.

Check before delivering

  • Objective said and shown in the first 30 s; recap matches the objective.
  • Every technical term spelled right in captions; captions <= 15 cps, never over code or the key part of a slide.
  • Same intro, colours and caption style as the previous modules.
  • qc.py passes; loudness consistent across modules (all at the same target).

Example requests

  • "Edit lesson 3 of my course: cut the mistakes, add step cards and captions."
  • "Split this 40-minute lecture into 5 lessons of about 8 minutes."
  • "Add a quiz question after the example and a recap card at the end."
Files (claude-video-editing-skills)
  • SKILL.md 7.1 KB
    ---
    name: course-module
    description: "Online course lesson editor for Claude: turns a teacher's recording (talking head, slides or screen) into a clean 3-10 min lesson with one learning objective, numbered steps, callouts, quiz cards, a recap, accessible captions (WCAG) and a transcript, consistent across every module. Use when the user says 'edit my course video', 'make a lesson', 'e-learning module', 'Udemy / Teachable lecture', 'training video for staff', 'add quiz cards and captions', or hands over lecture recordings. Runs on video-studio-core (transcribe, retakes, autocut, chapters, captions, deliver, QC)."
    license: MIT
    ---
    
    # Course Module
    
    One lesson, one objective, nothing the learner has to rewind to understand.
    
    - **Best for:** online courses, corporate training, onboarding lessons, university lectures, coaching programmes
    - **Inputs:** lesson recording (face, slides, screen or all three), slide deck export or screen capture, course name
    - **Default skin:** `whiteboard-host` · alternatives `tutorial-steps`, `highlighter-study`, `chapter-pro`
    - **Formats:** 16:9 lesson; optional 9:16 promo clip per module
    - **Durations:** 3-10 min per lesson (split longer recordings)
    - **Genre preset:** `--genre course-module`
    
    ## Before you start
    
    - Needs **video-studio-core** in the same skills folder. `CORE` = that folder, `SKILLS` = the skills folder.
    - First run: `python CORE/scripts/check_env.py`.
    - Ask for: the learning objective of this lesson (one sentence), terms and names to spell (`--prompt`), the course
      template (intro card, colours) so every module looks the same, and whether an LMS needs SCORM.
    - SCORM packaging is not done by the engine: deliver MP4 + VTT + transcript and let the LMS wrap them.
    - Accessibility is part of the job: captions on every lesson, transcript text, nothing conveyed by colour alone.
    
    ## How a pro cuts this
    
    **Structure (% of lesson)**
    
    | beat | % | what goes here |
    | --- | --- | --- |
    | objective | 8 | "By the end of this lesson you can..." on screen and said |
    | why it matters | 10 | the problem this solves for the learner |
    | teach | 55 | concept in 2-4 numbered steps, one idea per step |
    | example / practice | 17 | worked example, then a quiz card or exercise prompt |
    | recap + next lesson | 10 | 3 bullet recap, what comes next |
    
    **Pacing.** ASL 4-10 s; visual change every 8 s (callout, step card, slide change). Slower than YouTube: learners
    need time to read. Hold slides with text >= 1.5 s per line of text.
    
    **Cut rules.**
    - Keep the last good take (`--retakes`); remove fillers and long pauses but keep the teacher's rhythm.
    - Jump cuts are acceptable in lessons; hide the obvious ones with a slide or a punch-in.
    - Cut repetition and tangents; move asides to a "notes" document, not the video.
    - Split recordings at topic shifts into separate lessons (`story.py outline` shows topic shifts).
    
    **Transition grammar.** Cuts; a dissolve only at a scene change (talking head to slides). Same intro and outro card on
    every module.
    
    **Music and sound.** No music under teaching (bed at -32 dB only on the intro/outro card). Clear voice:
    `"eq": "clear"`, de-ess, -16 LUFS for LMS/web. No SFX except a soft tick on a quiz reveal.
    
    **Colour.** No look (accurate). Fix white balance and exposure only (auto correction).
    
    **Captions.** Accessible subtitles, sentence case, <= 15 cps, 2 lines max, high contrast (`"preset": "clean"`);
    always also deliver `.vtt`/`.srt` sidecars and a transcript.
    
    **What the client judges.** Clarity, consistency across modules, accessibility.
    
    **Common mistakes.** No stated objective; 25-minute lessons; reading slides aloud; music under explanation;
    captions too fast; every module with a different look; missing transcript.
    
    ## Make the edit
    
    ```bash
    python CORE/scripts/run.py lesson03.mp4 --style SKILLS/whiteboard-host --genre course-module --format 16:9 --smart-cuts --retakes --prompt "Kubernetes, kubectl, YAML" --title "Lesson 3: Deployments" --work m03 --stop-after plan
    python CORE/scripts/story.py outline --work m03 --genre course-module    # topic shifts -> split long recordings
    python CORE/scripts/render.py m03/plan.json --frames 2,30,90,180,280
    python CORE/scripts/render.py m03/plan.json
    python CORE/scripts/finish.py m03/plan.json --platform podcast
    ```
    
    `--platform podcast` gives -16 LUFS (LMS/web target). Slides or screen as a separate file: add `--clip slides.mp4`
    and place it with `clip` elements or full-frame `cutaways`. Split a long recording into lessons with `--cut` ranges
    or `story.py reorder --work m03 --order "s1-s24"` per lesson, then `run.py - --timeline ...` per lesson.
    
    ## Refine the plan
    
    - **Objective card:** `{"type": "titlecard", "kicker": "Lesson 3 of 8", "title": "Deployments", "subtitle": "Roll out a new version with zero downtime", "start": 0, "end": 4}`.
    - **Steps:** `{"type": "steps", "steps": [{"label": "Write the YAML", "at": 31}, {"label": "Apply", "at": 74}, {"label": "Verify", "at": 120}]}`
      or a `checklist` with `items[].at` when the teacher lists things.
    - **Callouts and highlights:** `{"type": "callout", "point": [880, 420], "x": 1450, "y": 220, "title": "replicas: 3"}`,
      `{"type": "text", "text": "rolling update", "fx": {"name": "highlight", "target": "rolling"}}`.
    - **Quiz card:** `{"type": "poll", "question": "Which command applies a manifest?", "options": [{"label": "kubectl apply", "pct": 100}, {"label": "kubectl run", "pct": 0}], "revealAt": 3}`
      (pause cue: hold 4-5 s), or a `card` with `title` and `items`.
    - **Recap:** `card` with 3 `items` timed to the teacher's recap words.
    - **Captions:** `{"type": "captions", "preset": "clean", "cpsMax": 15}`; fix terms in `words`.
    - **Audio:** `"audio": {"platform": "podcast", "voice": {"denoise": "light", "eq": "clear", "deess": true}}`; no `music`
      (or only on the intro card with `"fadeOut": 1`). Remove `sfx` the skin added.
    - **Colour:** `"color": {"auto": true}` only; no `look`.
    - Keep a `course_template.json` of the intro card, colours (`theme.vars`) and caption settings; paste it into every module.
    - `python CORE/scripts/editor.py review m03/plan.json --genre course-module`.
    
    ## Deliver
    
    ```bash
    python CORE/scripts/deliver.py m03/plan.json --to youtube,web-h265 --out-dir m03/deliver --burn no --cps 15 --chapters "0:00 Objective; 0:40 Why; 1:30 Steps; 5:10 Practice; 6:40 Recap"
    python CORE/scripts/qc.py m03/deliver/final_youtube.mp4 --plan m03/plan.json --platform youtube --strict
    ```
    
    Hand over per lesson: MP4, `captions.vtt` + `.srt`, a transcript (the caption text from `post.md` or the words),
    chapters. Name files `course_m03_deployments_v002.mp4`. SCORM: give the LMS these files; the engine makes no package.
    
    ## Check before delivering
    
    - Objective said and shown in the first 30 s; recap matches the objective.
    - Every technical term spelled right in captions; captions <= 15 cps, never over code or the key part of a slide.
    - Same intro, colours and caption style as the previous modules.
    - `qc.py` passes; loudness consistent across modules (all at the same target).
    
    ## Example requests
    
    - "Edit lesson 3 of my course: cut the mistakes, add step cards and captions."
    - "Split this 40-minute lecture into 5 lessons of about 8 minutes."
    - "Add a quiz question after the example and a recap card at the end."
    

Comments (0)

Sign in to join the conversation.

No comments yet.

Reviews (0)

No reviews yet.