ads-video

# Video Ad Script Writer > Portability note: triggers like `/ads video` and the `/ads strategy` orchestrator are OmegaOS slash-command conventions. The skill itself is environment-agnostic — invoke it directly with a product name/URL on any host; only the `/ads …` invocation phrasing is VPS-specific. ## Single-voice craft (do NOT parallelize) This is a **single-voice creative deliverable**, not a multi-angle analysis. The 9 scripts must read like one creative director wrote them in one sitting — consistent brand voice, recurring hook DNA, a coherent through-line. **Never fan out the script-writing across parallel subagents**: splitting the 9 scripts across workers fractures voice, duplicates research, and produces tonal whiplash between frameworks. Write all 9 in one pass, in one head. Allowed parallelism is **research-only** (Step 1): the `WebFetch` + `WebSearch` calls in Step 1 may run concurrently to gather context faster. The moment context is assembled, switch to single-threaded writing. **Craft upgrades (apply to every script):** 1. **Carry one hook DNA across all 9.** Pick the single sharpest angle for this product, then express it three ways (per framework) at three lengths — don't invent 9 unrelated concepts. The 30s and 60s should feel like the 15s with room to breathe, not different ads. 2. **Write voiceover for the ear, not the page.** Read every VO line aloud against the timecode — if it can't be spoken cleanly inside its slot at a natural pace (~2.5 words/sec), cut words until it can. Front-load the verb, kill filler ("basically", "literally", "in order to"). The word-count limits (15s: 30-40, 30s: 65-80, 60s: 130-160) are hard ceilings, not targets. 3. **Make on-screen text do its own job, not echo the VO.** Text overlays should add a stat, a price, or a punch the voiceover doesn't say — redundant text wastes the most valuable real estate on screen. Hook text must be readable in <0.8s (≤6 words, high contrast). ## Skill Purpose Generate complete, production-ready video ad scripts for three standard ad lengths: 15-second (TikTok/Reels/Bumper), 30-second (Pre-roll/Stories), and 60-second (YouTube/Facebook). Each script includes a shot-by-shot breakdown with exact timing, voiceover transcript, on-screen text overlays, B-roll suggestions, music/mood direction, and CTA placement. Scripts are written using three proven video ad frameworks and include a storyboard outline that a video editor or motion designer can follow without additional briefing. ## When to Use - User runs `/ads video <product>` - User asks for video ad scripts, video creative, or video content for ads - Called as a subagent from `/ads strategy` (the main orchestrator) - User needs scripts for TikTok ads, YouTube pre-roll, Instagram Reels, Facebook video ads, or Stories - User wants to brief a video editor or freelancer on ad creative ## Input Requirements - **Required:** Product or service to advertise (name, description, or URL) - **Optional:** Target audience persona, brand guidelines, existing footage/assets, desired tone, specific platform focus, offer details ## How to Execute ### Step 0 (optional): analyze reference ads with /watch If the user names competitor video ads or reference creatives (YouTube, TikTok, Instagram, X links, or local files), run /watch on 2 or 3 of them before Step 1. Each report.md documents the 0-10s hook (dense frames plus exact opening words), pacing (cuts per minute), and the full transcript. Feed that into the Video Brief: which hook patterns the winners use, how fast they cut, when the product appears. Write scripts that beat observed ads, not imagined ones. Skip when no references exist. ### Step 1: Gather Product & Brand Context If a URL is provided, fetch it using `WebFetch` and extract: - Product/service name and description - Key features and benefits - Price point and offer structure - Target audience indicators - Brand tone and visual style - Existing video content (YouTube channel, social media) - Testimonials or case studies - Competitor positioning Run supplementary research: ``` WebSearch: "[Product/Brand]" video ad WebSearch: "[Industry]" best video ad examples WebSearch: "[Industry]" TikTok ad examples viral WebSearch: "[Product category]" YouTube ad examples ``` Establish the video brief: ```markdown ## Video Brief **Product/Service:** [Name] **Business:** [Company name] **What it does:** [1-sentence description] **Who it's for:** [Target persona in 1 sentence] **Core benefit:** [The #1 reason someone buys this] **Key differentiator:** [What makes it different from alternatives] **Price/Offer:** [Current pricing, discount, free trial] **Desired action:** [What should the viewer DO after watching] **Tone:** [Professional / Casual / Energetic / Bold / Warm / Edgy / Playful] **Brand colors:** [Primary and secondary if known] **Existing assets:** [Footage, product photos, testimonials available] **Competitors to reference:** [Who they're competing against for attention] ``` ### Step 2: Apply Three Video Ad Frameworks Every script set uses all three frameworks. Each framework produces one version per ad length (15s, 30s, 60s) for a total of 9 scripts. --- #### Framework 1: Hook-Demo-CTA **Structure:** - **Hook (first 20-30% of runtime):** Bold opening statement, shocking stat, or provocative question that stops the scroll - **Demo (middle 50-60%):** Show the product in action. Prove the claim made in the hook. Use screen recordings, product shots, or demonstrations. - **CTA (final 10-20%):** Clear call to action with urgency or incentive. On-screen text + verbal CTA. **Best for:** SaaS demos, product showcases, tool walkthroughs, e-commerce product reveals **Psychology:** The hook creates attention, the demo creates desire through proof, the CTA creates action through urgency. --- #### Framework 2: Problem-Solution-Proof **Structure:** - **Problem (first 25-30%):** Show or describe the painful problem the audience faces. Make them feel the frustration. - **Solution (middle 40-50%):** Introduce the product as the answer. Show how it specifically solves the problem stated. - **Proof (final 20-30%):** Back it up. Testimonial clip, results screenshot, before/after, or stat. End with CTA. **Best for:** Pain-driven products, service businesses, health/wellness, financial services, any product that solves a clear problem **Psychology:** Pain activation (amygdala), then relief (dopamine release when solution is shown), then social validation (proof reduces risk). --- #### Framework 3: UGC-Style Testimonial **Structure:** - **Personal hook (first 15-25%):** Person speaking directly to camera with a relatable statement: "I tried everything for [problem] and nothing worked until..." - **Story/experience (middle 50-60%):** Tell the story of discovering and using the product. Show authentic, unpolished moments. Include specific details and results. - **Recommendation + CTA (final 15-25%):** Genuine endorsement with specific reasons why. End with "Link in bio" or "Check it out" style CTA. **Best for:** DTC brands, courses/coaching, beauty/wellness, apps, any product where social proof drives purchase **Psychology:** Mirror neurons fire when watching someone like us share an experience. UGC-style content bypasses ad skepticism because it looks like organic content. ### Step 3: Write Scripts for Each Length --- #### 15-Second Scripts (TikTok / Reels / Bumper Ads) Constraints: - Total runtime: 15 seconds - Word count (voiceover): 30-40 words max - On-screen text: 2-3 text cards max - One single focused message — no room for multiple points - Hook must land in first 1.5 seconds For each of the 3 frameworks, produce a script in this format: ```markdown ### 15s Script [Number] — [Framework Name] **Platform(s):** TikTok In-Feed, Instagram Reels, YouTube Bumper **Aspect ratio:** 9:16 (vertical) **Total runtime:** 15 seconds **Word count:** [count] words --- | Time | Visual / Shot | On-Screen Text | Voiceover / Audio | Music/SFX | |---|---|---|---|---| | 0:00-0:02 | [Shot description — camera angle, subject, action] | [Bold text overlay] | "[Spoken words]" | [Music mood: energetic opener / bass drop / trending sound] | | 0:02-0:05 | [Shot description] | [Text overlay if any] | "[Spoken words]" | [Music continues / transitions] | | 0:05-0:10 | [Shot description] | [Text overlay if any] | "[Spoken words]" | [Music mood] | | 0:10-0:13 | [Shot description] | [Text overlay if any] | "[Spoken words]" | [Music builds] | | 0:13-0:15 | [CTA shot — product shot, logo, URL] | [CTA text: "Shop Now" / "Link in bio" / "Try free"] | "[CTA spoken]" | [Music resolves / sting] | **B-Roll suggestions:** - [Suggestion 1 — specific, filmable scene] - [Suggestion 2] - [Suggestion 3] **Storyboard outline:** | Frame | Description | Notes | |---|---|---| | Frame 1 | [What the viewer sees — composition, subject, text position] | [Production note] | | Frame 2 | [description] | [note] | | Frame 3 | [description] | [note] | | Frame 4 | [description — CTA frame] | [note] | **Director's notes:** - [Pacing note — fast cuts, slow reveal, etc.] - [Tone note — energy level, facial expressions, urgency] - [Platform-specific note — TikTok native feel vs. polished for YouTube] ``` --- #### 30-Second Scripts (Pre-Roll / Stories / Standard Social) Constraints: - Total runtime: 30 seconds - Word count (voiceover): 65-80 words - On-screen text: 4-6 text cards - Hook must land in first 3 seconds - For YouTube pre-roll: hook before the 5-second skip button - Can include 1 testimonial quote or stat card For each of the 3 frameworks, produce a script in this format: ```markdown ### 30s Script [Number] — [Framework Name] **Platform(s):** YouTube Pre-Roll, Facebook Feed Video, Instagram Stories (multi-card) **Aspect ratio:** 16:9 (landscape for YouTube) / 9:16 (vertical for Stories) / 1:1 (square for Feed) **Total runtime:** 30 seconds **Word count:** [count] words --- | Time | Visual / Shot | On-Screen Text | Voiceover / Audio | Music/SFX | |---|---|---|---|---| | 0:00-0:03 | [Hook shot — must capture attention before skip] | [Bold hook text] | "[Hook line]" | [Attention-grabbing opener] | | 0:03-0:07 | [Shot description] | [Text overlay] | "[Spoken words]" | [Music mood] | | 0:07-0:12 | [Shot description — demo/problem visualization] | [Text overlay] | "[Spoken words]" | [Music transitions] | | 0:12-0:18 | [Shot description — solution/product in action] | [Key benefit text] | "[Spoken words]" | [Music builds] | | 0:18-0:24 | [Shot description — proof/testimonial/result] | [Proof text — stat or quote] | "[Spoken words]" | [Music continues] | | 0:24-0:28 | [CTA shot — product, offer, urgency element] | [Offer/CTA text] | "[CTA spoken]" | [Music resolves] | | 0:28-0:30 | [End card — logo, URL, CTA button] | [Brand + CTA] | "[Final line]" | [Music sting / outro] | **B-Roll suggestions:** - [Suggestion 1 — specific, filmable scene] - [Suggestion 2] - [Suggestion 3] - [Suggestion 4] - [Suggestion 5] **Storyboard outline:** | Frame | Time | Description | Camera | Text Position | |---|---|---|---|---| | Frame 1 | 0:00-0:03 | [description] | [angle/movement] | [where text appears] | | Frame 2 | 0:03-0:07 | [description] | [angle] | [text position] | | Frame 3 | 0:07-0:12 | [description] | [angle] | [text position] | | Frame 4 | 0:12-0:18 | [description] | [angle] | [text position] | | Frame 5 | 0:18-0:24 | [description] | [angle] | [text position] | | Frame 6 | 0:24-0:30 | [CTA card] | [static/zoom] | [centered] | **Director's notes:** - [Pacing and energy direction] - [Transition style between shots — cut, dissolve, swipe, zoom] - [Color grading / visual mood direction] - [Audio mixing notes — music vs. voiceover balance] ``` --- #### 60-Second Scripts (YouTube / Facebook / Long-Form Social) Constraints: - Total runtime: 60 seconds - Word count (voiceover): 130-160 words - On-screen text: 6-10 text cards - Full narrative arc — beginning, middle, end - Room for 1-2 testimonial clips, multiple product shots, and a strong CTA sequence - Hook must still land in first 3 seconds For each of the 3 frameworks, produce a script in this format: ```markdown ### 60s Script [Number] — [Framework Name] **Platform(s):** YouTube In-Stream, Facebook Feed Video, LinkedIn Video **Aspect ratio:** 16:9 (primary), adaptable to 9:16 and 1:1 **Total runtime:** 60 seconds **Word count:** [count] words --- | Time | Visual / Shot | On-Screen Text | Voiceover / Audio | Music/SFX | |---|---|---|---|---| | 0:00-0:03 | [Hook shot] | [Hook text] | "[Hook line]" | [Attention opener] | | 0:03-0:08 | [Context/problem setup] | [Supporting text] | "[Lines]" | [Music mood] | | 0:08-0:15 | [Problem deepens / agitate] | [Pain point text] | "[Lines]" | [Music tension builds] | | 0:15-0:22 | [Solution introduction] | [Product/brand reveal text] | "[Lines]" | [Music shift — hopeful] | | 0:22-0:32 | [Demo / how it works / key features] | [Feature/benefit callouts] | "[Lines]" | [Music upbeat] | | 0:32-0:42 | [Proof — testimonial, results, case study] | [Proof stat or quote] | "[Lines]" | [Music maintains] | | 0:42-0:50 | [Additional benefit or second proof point] | [Secondary benefit text] | "[Lines]" | [Music builds to climax] | | 0:50-0:55 | [CTA — offer reveal, urgency] | [Offer text + CTA] | "[CTA spoken]" | [Music peaks] | | 0:55-0:60 | [End card — logo, URL, tagline] | [Brand + final CTA] | "[Closing line]" | [Music resolves] | **B-Roll suggestions:** - [Suggestion 1 — opening establishing shot] - [Suggestion 2 — problem visualization] - [Suggestion 3 — product detail shots] - [Suggestion 4 — customer reaction / testimonial setting] - [Suggestion 5 — lifestyle / aspirational imagery]