.claude/skills/class-script-writer/SKILL.md
Turn a raw idea, source tweet, video transcript, feature announcement, or rough outline into a polished video script for Ray's Agentic Coding School classes. Use this skill whenever the user says "write a class script", "turn this into a class video", "script this for the [business/claude-code/prompt-engineering/skills/context-engineering/techniques/claude-chat/claude-cowork/correction] class", "draft an agentic coding school video", or whenever the user is staging content in `/Users/ray/Desktop/ray-os/projects/agentic-coding-school/to-film/` and wants it written up. Default to this skill for any new file destined for the `to-film/` tree, even if the user does not name it explicitly. Produces an essay-style markdown script with the correct frontmatter, horizontal rules between sections, inline image embed placeholders, and a Demo section, written in Ray's direct second-person voice.
npx skillsauth add theramjad/ray-os class-script-writerInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Write video scripts for Ray's Agentic Coding School in the same voice and structure as the already-filmed classes. The output is not an outline or a bullet list of beats. It is an essay the camera follows, with clean section breaks, embedded image placeholders, and a concrete demo at the end.
Trigger on any of these:
/Users/ray/Desktop/ray-os/projects/agentic-coding-school/to-film/<class>/ and wants it written, or they hand you a source (tweet, transcript, outline, announcement) and mention filming, classes, or the school.to-film/ and asks to rewrite it in the style of an existing filmed class. Overwrite the file in place.If the class folder is ambiguous, ask which class it belongs to before writing. The classes are:
business, claude-chat, claude-code, claude-cowork, context-engineering, correction, prompt-engineering, skills, techniques.
Gather before writing. Ask only for what is missing.
"10-14 min" unless the user specifies.The output path is /Users/ray/Desktop/ray-os/projects/agentic-coding-school/to-film/<class>/<slug>.md.
Every script starts with this block. Fill every field.
---
duration: "10-14 min"
batch: 1
order: 2
batch_name: "Research & Intelligence"
class: "business"
chapter: "Research & Intelligence"
---
Pick batch and order by scanning the sibling files in the target class folder. If everything there is order 1, use 2. Do not leave any field blank.
Read references/style-guide.md for the full rules. The short version:
## headings, never #. Separate major sections with --- on their own line.Source: https://... right under the claim, not bundled at the top.The reference scripts to imitate are bundled in references/:
references/scaling-taste.md — long-form, idea-dense, multiple "Why this works" and "The X Loop" sections. Good template for conceptual videos.references/progressive-disclosure.md — tighter, problem-to-solution arc with bold one-liners and industry context. Good template for a technique or pattern video.Skim both before writing. Match whichever matches the user's topic best.
Every script has these, in this order. Section names can flex to fit the topic.
A cold-open section with a one-sentence thesis. No fluff intro. State what this video is about in the first paragraph.
The problem. What is broken today, or what everyone is doing wrong. One image placeholder usually lands here.
The core insight or solution. The single idea the video exists to deliver.
Supporting sections. Two to four of them. How it works, why it works, where it came from, what the layers are. Use --- between each.
Demo. A concrete, numbered walkthrough of what the camera will show. Not vague. Specific tools, specific outputs.
Key Insight (optional but recommended). A single blockquote that could be pulled as a tweet. Format:
> A single crisp sentence that captures the whole video.
A closing beat. One to three lines. What changes for the viewer after they watch this.
Use Obsidian embed syntax. The images directory is keyed on the slug.
![[images/<slug>/<name>.png]]
Place two to four image embeds through the script, typically one after the problem section, one after the core insight, and one or two in the supporting sections. Describe what each image should show inline in square brackets above the embed so the excalidraw-gen skill can pick them up later:
[IMAGE: dark chalkboard, two panels labeled Knowledge and Your Life, lightning bolt between them]
![[images/notebooklm-intake/the-problem.png]]
references/ to lock in voice.batch, order, and avoid duplicate chapters.— and –. Replace them with commas, periods, or rewritten sentences. This is a hard rule, not a preference./Users/ray/Desktop/ray-os/projects/agentic-coding-school/to-film/<class>/<slug>.md.excalidraw-gen skill if the user wants them.deep-research-with-exa.md are outlines, not scripts. Do not copy that shape. Copy the scaling-taste and progressive-disclosure shape instead.# Title at the top of the file.Ray has filmed roughly a hundred of these videos. The ones that land have the same shape: a thesis in the first paragraph, a crisp problem, an insight that feels earned, a demo the viewer can picture, a closing line that makes them want to try it today. The ones that flop are outlines dressed up as scripts, or model-voiced essays with em dashes and hedged claims.
The job of this skill is to hit that shape every time, in Ray's voice, with the right file path and frontmatter, so the script is ready to film without a rewrite pass.
content-media
Edit raw screen recording videos with synced audio. Removes repeated takes, false starts, silence, filler, and [MUSIC] tags to produce a clean final cut. Use when the user provides a raw/unedited MP4 video file and wants it edited down. Triggers on: "edit this video", "cut this video", providing a raw MP4 for editing, or any request to remove bad takes from a recording.
tools
Monitor Twitter/X for trending AI and developer tool discussions, score them, and deliver pre-digested content briefs. Use this skill whenever the user wants to find trending topics on Twitter, check what people are talking about in AI/tech, scout for video ideas from Twitter, or says things like "what's trending", "anything new on Twitter", "find me something to talk about", "scout Twitter", or "what are people saying about X". Also triggers on "monitor layer", "input layer", or "fountainhead" when Twitter is relevant.
development
# /triage — Inbox Triage ## Description Scan all connected communication channels, prioritize items by urgency, and draft responses in your voice. Clear your inbox in minutes. ## Arguments - `quick` — Tier 1 items only, no drafts (fastest) - `digest` — Full scan with summaries, drafts for Tier 1-2 - (no argument) — Full scan with drafts for everything actionable ## Instructions You are running inbox triage for {{YOUR_NAME}}. The goal is to process all incoming messages quickly and surface wh
tools
YouTube data utility — fetch transcripts, metadata, thumbnails, and search for videos. Wraps the Supadata API and yt-dlp into a single CLI tool. Use whenever you need YouTube data like transcripts, video metadata, thumbnail images, or need to search YouTube for videos matching a query. Triggers on any YouTube data fetching need from other skills or direct user requests.