doany-ai

Generating broadcast-quality video from text requires precise control over motion, audio, and aspect ratio. Wan 2.7 on RunComfy provides multi-reference conditioning, audio-driven lip-sync, and clean exit codes for production workflows.

Videos shot in one aspect ratio often need to be repurposed for different platforms without reshooting. This skill enables AI agents to extend video canvases spatially using the RunComfy CLI, preserving central action while adding matching environment beyond the original frame.

Removing unwanted objects, watermarks, or background distractions from video clips is tedious with traditional editors. This skill routes your request to the right RunComfy inpainting model so you can clean up video frames with a single prompt.

Video clips have per-call duration limits. This skill extends existing Veo clips by chaining continuations with consistent motion, lighting, and subject identity for longer narratives.

Choosing the wrong AI video model wastes time and produces poor results. This skill reads the user intent and selects the best model from the RunComfy catalog with correct prompting patterns.

Creating professional video content requires expensive equipment and expertise. Seedance 2.0 Pro on RunComfy enables AI-powered cinematic video generation with native lip-sync, multi-modal references, and production-ready output in seconds.

Calling AI models from Claude Code requires handling authentication, polling, and file downloads. This skill teaches Claude Code to use the runcomfy CLI for seamless access to hundreds of AI models including image generation, video, and face swapping.

Change the lighting direction, color temperature, and mood of any still image without redoing the shot. This skill routes to purpose-built AI relighting models for professional results.

Need to edit product photos, swap backgrounds, or batch process images? Nano Banana 2 Edit on RunComfy offers batch support for up to 20 images with strong subject identity preservation and spatial edit controls.

Quickly generate images with Google Nano Banana 2 via RunComfy CLI. Includes optimized prompting patterns, batch generation support, and web-grounded context for current events.

Creating lip-synced video requires choosing the right AI model for your specific use case. This skill automatically routes your request to the best RunComfy endpoint — Sync Labs, OmniHuman, or Kling — based on your input format and quality needs.

Kling 3.0 is a multi-shot video model with synchronized audio, but navigating six endpoints across three tiers is complex. This skill routes your request to the right Kling 3.0 endpoint and invokes the RunComfy CLI.

Choosing the wrong image-to-video model wastes iterations and credits. This skill routes your request to the best RunComfy model for your goal — portrait animation, lip-sync, or multi-modal composition — so you get quality output faster.

Need to expand an image beyond its original canvas or change aspect ratio? This skill uses RunComfy's AI models to extend images while preserving identity and style.

Mask-driven image inpainting removes unwanted objects, replaces backgrounds, and fills masked regions using AI models. This skill routes your edits to the optimal RunComfy model for precise or description-based results.

Choosing the right AI model for each image edit wastes time and iterations. This skill reads your intent and routes to the best RunComfy model automatically, with documented prompting patterns for each route.

Creating video content requires expensive software and technical expertise. HappyHorse 1.0 generates high-quality 1080p videos with synchronized audio from text prompts, making professional video creation accessible to anyone with a RunComfy account.

Naive prompts against GPT Image 2 drift on identity, brand, and multilingual text. This skill gives Claude, Codex, and Claude Code the documented preservation language, multi-reference numbering, and script-quoting patterns that produce sharp, on-brief edits on the first try.

Creating images with embedded text, logos, and multilingual typography is unreliable with most models. This skill routes requests to GPT Image 2 via RunComfy CLI for directive-precise image generation and editing.

Precise local image edits often fail because prompts are not shaped for the model. This skill bundles Flux 1 Kontext Pro's documented prompting patterns so edits come out sharp on the first try. It also routes you to the right sibling model when Kontext is not the best fit.

Getting high-quality AI-generated images often requires complex setup and slow iteration cycles. This skill connects you to Black Forest Labs' Flux 2 Klein through the RunComfy CLI for sub-second image generation with optimized prompting patterns.

Face swapping requires choosing between multiple AI models for different use cases. This skill routes requests to the right model based on whether you need video or still images, single or batch processing, and identity or motion preservation.

Create studio-quality songs and instrumental tracks from text descriptions without manual music production expertise. ElevenLabs Music generates 44.1kHz stereo audio with section-level control using structured lyrics and style prompts.

Choosing the right pose-conditioned model is confusing when video motion transfer, still-image pose locking, and multi-condition ControlNet stacks each need different endpoints. This skill routes your request to the correct RunComfy Model API endpoint automatically.