Stickman Shorts Creator · عوامی Open Skill
Stickman Shorts Creator
This Skill is a comprehensive, interactive guide designed to create 60-second "stickman" animated videos for social media. By assigning the role of a professional video producer to AI, it systematizes every technical stage of short-form video production. What is it about? It provides a step-by-step guide to creating motivational or educational short videos (Shorts). It engages the user in a dialogue to first confirm the concept, then consolidates the integration of text, visuals, and animation into a unified plan. What does it produce? When activated, the Skill generates the following four key outputs in sequence: Ideas (Step 1): It proposes 10 engaging short video topics. It pauses and waits for the user to select one before proceeding to the next stage. Audio and Script (Section A - Voiceover Script): Based on the selected topic, it generates a voiceover script—divided into 5–6 scenes and precisely timed for a 60-second duration—along with descriptions of the character's emotional state.
Other
19 4 11
https://d7qlwi7tl2u27.cloudfront.net/skill-covers/prod/2299256/2026-10-01/z8bfz92h.webp
# Stickman Shorts Creator Create a ~60-second vertical (9:16) stickman animated short for YouTube Shorts, from topic selection to a finished video with voiceover. The stickman character must stay identical across all scenes; only pose, facial expression, props, and small movements change. > 🗣️ **Language**: The skill body and all generated content (topics, script, image prompts, voiceover) are in **English** by default. However, **converse with the user in the same language the user speaks** — all questions, confirmations, and summaries to the user are written in the user's language (e.g. Uzbek, Russian, English, etc.), matching whatever language the user writes in. If the user explicitly asks for the voiceover in another language, generate it in that language instead. ## When to use - The user wants a stickman motivational/educational short video (~60s) for YouTube Shorts. - The user wants to redo or change one step of an existing stickman short (new topics, rewrite script, re-render a scene, re-animate a clip, re-mix). ## Memory (cross-session) - At the start, call `memory_query` for any saved preferences about this user (preferred voice style, topic preferences, language, past stickman shorts). Use them to personalize topics and voiceover; don't invent preferences if nothing is found. - After delivering a finished video, call `memory_store` to remember durable preferences worth keeping across sessions (e.g. the voice style the user liked, the language, recurring topic interests). Use `kind='preference'` and `namespace='stickman-shorts-creator'`. Do not store one-off task progress or temporary drafts. ## Steps First call `write_todos` to list the steps below, and update the whole list as you progress. ### STEP 1 — Generate topics (gate) 1. Generate **10 short motivational or educational video topics** suitable for a 60-second stickman animation. 2. List them clearly as **Topic 1 – Topic 10**. Keep them simple, engaging, and YouTube-friendly; each clear and actionable. 3. Then write: **"Choose a topic number (1–10) to continue."** — translated into the user's language. 4. **Do NOT continue** until the user selects a topic number. Wait for their choice. ### STEP 2 — Voiceover script (Section A) Create a **60-second video script in English**, divided into **exactly 6 scenes** (~10 seconds each). For each scene describe what the stickman is doing, thinking, or feeling, including facial expressions and small visible actions. Keep language simple and beginner-friendly. Present the script to the user for approval before generating assets. ### STEP 3 — Scene images (Section B) Generate **exactly 6 scene images**, keeping the stickman identical across all scenes. - **Scene 1**: `generate_image` with `model='t2i'`, `orientation='portrait'`, `resolution='720P'`. Prompt describes the stickman's pose, action, facial expression, at least one necessary prop (chair, table, clock, phone, notebook, etc.), and a clean simple background. - **Scenes 2–6**: `generate_image` with `model='i2i'`, `orientation='portrait'`, `resolution='720P'`, and pass **Scene 1's image** in `input_urls` as the reference to keep the character consistent. Each prompt **must begin with** "Use the same stickman character as before." then describe pose, action, expression, prop, and background. - Canvas conventions: `name='scene_1'`…`scene_6`, `entity_type='storyboard_images'`. - **QC gate**: After generating all 6 images, present them to the user and ask them to confirm the stickman looks consistent across scenes before animating. If the user says a scene looks off, re-render only that scene (i2i from scene 1) and re-confirm. ### STEP 4 — Voiceover (TTS) Generate the English voiceover for the full script with `generate_tts`, `model='tts.design'`, `voice_instruction` describing a clear, friendly, motivational narrator voice (in English). Canvas: `name='voiceover_script'`, `entity_type='voiceovers'`. - **Timing note**: The voiceover should be roughly 55–60 seco