Purpose
Turn a show format and an episode topic into a shootable segment-by-segment outline — with per-segment runtime, host direction, transition cues, and platform-tuned hooks — so the production team walks into the shoot or edit with a complete structure instead of a blank page.
When to Use
Invoke this skill when the user asks to:
- Outline or script an episode of a show or series
- Break a topic down into segments with timing
- Plan a shoot or recording session for episodic content
- Convert a rough episode idea into a shootable structure
- Build a repeatable episode template for a recurring format
- Add host direction, hooks, or transition notes to an existing rundown
Inputs Required
Before starting, collect:
- Show format: Interview / solo monologue / panel / video essay / reaction / tutorial / narrative / hybrid
- Episode topic: The specific subject, story, or question this episode covers
- Target runtime: Total episode length in minutes
- Primary platform: YouTube / TikTok / Instagram Reels / podcast / long-form streaming / multi-platform
- Host(s) and role: Who is on camera/mic and their function (main host, co-host, guest, narrator)
- Episode goal: What the viewer should feel, learn, or do by the end
- Recurring segments (if any): Any fixed bits, sponsor reads, or callbacks the format always includes
If runtime or platform is missing, default to a 12-minute YouTube episode and note the assumption.
Instructions
Step 1: Confirm the Format Spine
Map the episode to a proven structure for its format type:
- Interview: Cold open hook → intro/context → 3–5 topic blocks → signature closing question → CTA
- Solo monologue / video essay: Hook → thesis → 3–4 argument beats → counterpoint/turn → payoff → CTA
- Tutorial: Hook (result preview) → what you'll need → sequential steps → common mistakes → recap → CTA
- Panel: Hook → guest intros → 2–4 debate prompts → hot-take round → closing statements → CTA
- Narrative: Cold open → inciting question → escalating beats → climax → resolution → CTA
Select the spine and state it in one line before building segments.
Step 2: Build the Segment-by-Segment Outline
For every segment, produce a block containing:
- Segment name + number
- Runtime: Duration in mm:ss and a running cumulative timestamp (must sum to target runtime)
- Purpose: One sentence — why this segment exists in the episode
- Content beats: 2–4 bullet points of what actually gets said or shown (specific to the topic, not generic)
- Host direction: Delivery, energy, blocking, or camera note (e.g., "drop energy, lean in — this is the confession beat")
- Transition cue: The exact line, gesture, or edit that moves into the next segment
Allocate runtime deliberately: the hook gets 5–10% of total, the core value segments get 60–70%, the close and CTA get the remainder. Never let a single mid-episode segment exceed 25% of total runtime without a subdivided beat structure.
Step 3: Engineer the Platform-Specific Hooks
Write hooks tuned to how the primary platform surfaces retention:
- YouTube: A 0–15s hook that names the payoff and creates an open loop; note a mid-roll re-hook around the 40% mark to survive the drop-off cliff
- TikTok / Reels: A 0–3s pattern interrupt (visual or verbal) plus a "watch till the end" tension line in the first 8 seconds
- Podcast: A cold-open teaser clip pulled from the strongest later moment, then a spoken hook question
- Long-form streaming: A cold open that establishes stakes before any branding
Deliver 2 hook variants for the opener so the producer can A/B or choose on the shoot day.
Step 4: Add Production Notes and CTA
- Flag any B-roll, graphics, or lower-third needs inline per segment
- Insert sponsor read / recurring segment placement if provided in inputs
- Write a specific CTA matched to the episode goal (subscribe, comment prompt, next-episode tease, link in bio) — not a generic "like and subscribe"
Output Format
- Structure: Numbered segment blocks in sequence, each with the six sub-fields from Step 2; a hook section up top with 2 variants; a running timing column
- Length: 700–1,100 words depending on runtime and segment count
- Format: Structured rundown — labeled segment blocks with bolded field names, scannable, not prose paragraphs
- Sections: Format spine (1 line) → Opening hooks (2 variants) → Segment-by-segment outline → Production notes summary → CTA
- Tone: Working showrunner — decisive, production-ready, written to be handed to a host or editor as-is
Quality Checklist
Before delivering output, verify:
- [ ] Segment runtimes sum exactly to the target total (show the cumulative timestamps)
- [ ] Every segment has all six fields: name, runtime, purpose, content beats, host direction, transition cue
- [ ] Content beats reference the actual episode topic — no placeholder or generic filler
- [ ] The hook section contains 2 distinct opener variants tuned to the named platform
- [ ] No single mid-episode segment exceeds 25% of total runtime
- [ ] The CTA is specific to the episode goal, not a boilerplate subscribe line
Notes
- The cold open earns the rest of the episode — spend disproportionate care on the first 15 seconds regardless of format
- If the topic is thin for the target runtime, recommend cutting runtime rather than padding segments — pacing beats length on every platform
- For recurring shows, this outline doubles as a reusable template: lock the format spine and swap only the topic-specific content beats each episode
How to use this skill
This file is a Claude Code skill in Anthropic's standard format — markdown with YAML frontmatter at the top.
Installing in Claude Code:
~/.claude/skills/<folder-name>/SKILL.md— auto-loads in every session anywhere on your machine (recommended for skills you'll use across multiple projects)<your-project>/.claude/skills/<folder-name>/SKILL.md— auto-loads only when working in that project (recommended for project-specific skills)
Restart Claude Code or open a new session to register.
Using in Claude.ai, ChatGPT, Gemini, or other AI tools:
- Drag this file into a new chat, or copy and paste its contents
- Send your task or request as your next message — the AI will apply the skill where it fits
For Claude Code skill details: https://docs.anthropic.com/en/docs/claude-code/skills