Sensual Video Lite (SVL) · skill pública Open Skill
Sensual Video Lite (SVL)
🚀 **The platform favorite is back—cleaner, smarter, and more consistent.** Sensual Video Lite transforms an uploaded or tagged image into identity-preserving edits and immersive videos, paired with interactive character chat. The character remembers her appearance, mood, pose, outfit, and current scene, allowing every new moment to continue naturally. ## ✨ Features | Feature | What it does | |---|---| | 🔒 **Identity Preservation** | Maintains the character’s face, body build, hair, lighting, color grade, and visual style. | | 🏷️ **Tagged Asset Support** | Tag any image to use it as the source for the next generation. | | 🧠 **Stateful Memory** | Tracks the current pose, facing direction, gaze, outfit, scene, contact, and previous action. | | 🎯 **Smart Intent Builder** | Converts short requests into detailed prompts with movement, camera, continuity, and end-state planning. | | 👁️ **Perspective Detection** | Automatically distinguishes first-person interactions from character-led actions. | | 📸 **Adaptive Image Editing** | Selects the most suitable image-editing model based on the request. | | 🎥 **Source-Aware Video** | Creates 10-second I2V clips by default while preserving the source image’s aspect ratio. | | 🗣️ **Native Character Dialogue** | Generates mood-appropriate video dialogue while preventing prompt text from being spoken. | | ⚖️ **Advanced Physics** | Improves body weight, balance, contact, soft-tissue response, hair movement, fabric behavior, and pose transitions. | | 💬 **Two Experience Modes** | Use Scenes for quick generations or Sensual Chat for an ongoing character experience. | | 💫 **Customizable Personality** | Choose Bold, Flirty, Angry, Arrogant, Loving, Caring, Possessive, Obsessed, or type a custom mood. | | ▶️ **Context-Aware Buttons** | Receives fresh generation options based on the latest message, pose, outfit, scene, and completed action. | | 👥 **Two-Character Support** | Tracks separate identities, moods, voices, poses, outfits, and roles for both characters. | | 🔄 **Continuity Handling** | Continues from the latest successful image while preserving unmodified scene and character details. | | 🖼️ **Clean Inline Images** | Displays generated images directly inside the conversation. | | 🚫 **No Forced Processing** | Avoids automatic cropping, aspect changes, and unnecessary upscaling. | ## 📋 How a Session Works 1. Choose **Scenes** or **Sensual Chat**. 2. Upload or tag an image. 3. Name the character or skip naming. 4. Select a mood. 5. Talk, tap a suggested action, or describe what happens next. 6. Continue from the latest image or turn it into a video. The character stays in role while the skill silently handles identity, continuity, model selection, motion planning, physics, and prompt building. 🎥 Optimized for natural movement, reliable identity preservation, expressive character behavior, and high-quality image-to-video generation. Built for everyone who loved the original and wanted a cleaner, more intelligent experience that still feels unmistakably like the same character.
Film
4345 118 668
https://d7qlwi7tl2u27.cloudfront.net/skill-covers/prod/2693085/2026-10-01/kykpv7eg.jpg