미디어 스킬

Mask-driven image inpainting on RunComfy via the `runcomfy` CLI. Routes to Tongyi MAI Z-Image Turbo Inpainting (the dedicated inpainting endpoint with mask, strength, and control-scale) and to identity-preserving edit models (Nano Banana 2 Edit, GPT Image 2 Edit, FLUX Kontext Pro) when a mask isn't available and the region must be described instead. Use for object removal, watermark removal, region replacement, blemish cleanup, and any controlled local edit where a binary mask defines the...

Mask-driven image inpainting on RunComfy via the `runcomfy` CLI. Routes to Tongyi MAI Z-Image Turbo Inpainting (the dedicated inpainting endpoint with mask, strength, and control-scale) and to identity-preserving edit models (Nano Banana 2 Edit, GPT Image 2 Edit, FLUX Kontext Pro) when a mask isn't available and the region must be described instead. Use for object removal, watermark removal, region replacement, blemish cleanup, and any controlled local edit where a binary mask defines the...

Image outpainting on RunComfy via the `runcomfy` CLI — extend a still beyond its original canvas, fill in what the camera didn't capture, change aspect ratio (square → 16:9, portrait → landscape) while preserving the original content. Routes across Nano Banana 2 Edit (default, spatial-language driven), GPT Image 2 Edit (multi-ref with reference-style matching), FLUX Kontext Pro (single-shot maximum-preservation), and the brand edit endpoints (Seedream / Dreamina / Qwen / FLUX 2). Picks the...

Image outpainting on RunComfy via the `runcomfy` CLI — extend a still beyond its original canvas, fill in what the camera didn't capture, change aspect ratio (square → 16:9, portrait → landscape) while preserving the original content. Routes across Nano Banana 2 Edit (default, spatial-language driven), GPT Image 2 Edit (multi-ref with reference-style matching), FLUX Kontext Pro (single-shot maximum-preservation), and the brand edit endpoints (Seedream / Dreamina / Qwen / FLUX 2). Picks the...

Image outpainting on RunComfy via the `runcomfy` CLI — extend a still beyond its original canvas, fill in what the camera didn't capture, change aspect ratio (square → 16:9, portrait → landscape) while preserving the original content. Routes across Nano Banana 2 Edit (default, spatial-language driven), GPT Image 2 Edit (multi-ref with reference-style matching), FLUX Kontext Pro (single-shot maximum-preservation), and the brand edit endpoints (Seedream / Dreamina / Qwen / FLUX 2). Picks the...

정지 영상을 비디오로 변환하는 가이드: 모델 선택, 모션 프롬프팅, 카메라 움직임. Wan 2.5 i2v, Seedance, Fabric, Grok Video를 각각 사용해야 하는 경우를 다룹니다. 용도: 이미지 애니메이션, 정지 영상으로 비디오 생성, 모션 추가, 제품 애니메이션. 트리거: image to video, i2v, animate image, still to video, add motion to image, image animation, photo to video, animate still, wan i2v, image2video, bring image to life, animate photo, motion from image

We need to translate the given text from English to Korean. The text describes a skill called "image-to-video" but the instruction says not to include the name unless it appears in the source text. The name "image-to-video" does appear in the source? Actually the source says "Animate any still image on RunComfy — this skill is a smart router..." The skill name is not explicitly given as "image-to-video" in the text; the directory item type is "agent skill" and the name to preserve is "image-to-video". But the instruction says "Do not include the name unless it appears in the source text." The name "image-to-video" does not appear in the source text. So we should not add it. We just translate the text inside <text>. The text: "Animate any still image on RunComfy — this skill is a smart router that matches the user's intent to the right i2v model in the RunComfy catalog. Picks HappyHorse 1.0 I2V (A

We need to translate the given text from English to Korean. The text describes a skill called "image-to-video" but the instruction says not to include the name unless it appears in the source text. The name "image-to-video" appears in the source? Actually the source says "Animate any still image on RunComfy — this skill is a smart router..." The name is not explicitly given as "image-to-video" in the text? Wait, the directory item type is "agent skill" and the name to preserve is "image-to-video". But the instruction says: "Do not include the name unless it appears in the source text." The source text does not contain the string "image-to-video". So we should not add it. We just translate the text inside <text>. Also preserve product names, protocol names, URLs, numbers, technical terms. So "RunComfy", "HappyHorse 1.0 I2V", "Arena #1", "Wan 2.7", "audio_url", "Seedance

Animate any still image on RunComfy — this skill is a smart router that matches the user's intent to the right i2v model in the RunComfy catalog. Picks HappyHorse 1.0 I2V (Arena #1, native audio, identity preservation) for general animations, Wan 2.7 with `audio_url` for custom-voiceover lip-sync, or Seedance 2.0 Pro for multi-modal animation from image + reference video + reference audio. Bundles each model's documented prompting patterns so the caller gets sharper output without burning...

정지 영상을 비디오로 변환하는 가이드: 모델 선택, 모션 프롬프팅, 카메라 움직임. Wan 2.5 i2v, Seedance, Fabric, Grok Video를 각각 사용해야 할 시점과 함께 다룹니다. 용도: 이미지 애니메이션, 정지 영상으로 비디오 만들기, 모션 추가, 제품 애니메이션. 트리거: image to video, i2v, animate image, still to video, add motion to image, image animation, photo to video, animate still, wan i2v, image2video, bring image to life, animate photo, motion from image

We need to translate the given text from English to Korean. The text describes a skill called "image-to-video" but the instruction says not to include the name unless it appears in the source text. The name "image-to-video" does appear in the source? Actually the source text starts with "Animate any still image on RunComfy — this skill is a smart router..." The name "image-to-video" is not explicitly in the source text; it's given as the directory item type name. So we should not include it. The instruction says: "Do not include the name unless it appears in the source text." It does not appear. So we translate only the text inside <text>. We need to preserve product names, protocol names, URLs, numbers, technical terms. So "RunComfy", "HappyHorse 1.0 I2V", "Arena #1", "Wan 2.7", "audio_url", "Seedance 2.0 Pro" should remain as is. Also "i2v" is a

RunComfy에서의 Kling 3.0 비디오 생성. Kling 3.0(또는 Kling V3.0)은 Kuaishou Technology의 3세대 멀티샷 비디오 모델로, 기본 동기화 오디오와 샷 간 일관된 캐릭터 정체성을 제공합니다. 이 스킬은 세 가지 렌더링 등급(Standard, Pro, 4K)과 두 가지 모드(텍스트-투-비디오, 이미지-투-비디오)에 걸친 모든 6개의 Kling 3.0 엔드포인트를 다룹니다. 로컬 RunComfy CLI를 통해 runcomfy run kling/kling-3.0/ / 을 호출합니다. "kling", "kling 3.0", "kling v3", "kling pro" 등에서 트리거됩니다.

RunComfy에서의 Kling 3.0 비디오 생성. Kling 3.0(또는 Kling V3.0)은 Kuaishou Technology의 3세대 멀티샷 비디오 모델로, 기본 동기화 오디오와 샷 간 일관된 캐릭터 정체성을 제공합니다. 이 스킬은 세 가지 렌더링 등급(Standard, Pro, 4K)과 두 가지 모드(텍스트-비디오, 이미지-비디오)에 걸친 모든 6개의 Kling 3.0 엔드포인트를 다룹니다. 로컬 RunComfy CLI를 통해 runcomfy run kling/kling-3.0/ /을 호출합니다. "kling", "kling 3.0", "kling v3", "kling pro" 등에서 트리거됩니다.

Kling 3.0 RunComfy 비디오 생성. Kling 3.0(일명 Kling V3.0)은 Kuaishou Technology의 3세대 멀티샷 비디오 모델로, 기본 동기화 오디오와 샷 간 일관된 캐릭터 정체성을 갖추고 있습니다. 이 스킬은 세 가지 렌더링 등급(Standard, Pro, 4K)과 두 가지 모드(텍스트-투-비디오, 이미지-투-비디오)에 걸친 6가지 Kling 3.0 엔드포인트를 모두 다룹니다. 로컬 RunComfy CLI를 통해 runcomfy run kling/kling-3.0/ / 호출을 실행합니다. "kling", "kling 3.0", "kling v3", "kling pro" 등에서 트리거됩니다.

RunComfy에서의 Kling 3.0 비디오 생성. Kling 3.0(또는 Kling V3.0)은 Kuaishou Technology의 3세대 멀티샷 비디오 모델로, 기본 동기화 오디오와 샷 간 일관된 캐릭터 정체성을 제공합니다. 이 스킬은 세 가지 렌더링 등급(Standard, Pro, 4K)과 두 가지 모드(텍스트-투-비디오, 이미지-투-비디오)에 걸친 모든 6개의 Kling 3.0 엔드포인트를 다룹니다. 로컬 RunComfy CLI를 통해 runcomfy run kling/kling-3.0/ /을 호출합니다. "kling", "kling 3.0", "kling v3", "kling pro"에서 트리거됩니다.

We need to translate the given English text into Korean, preserving the specified names: lipsync, RunComfy, runcomfy, ByteDance OmniHuman, Sync Labs sync v2 / Pro, Kling lipsync, Creatify lipsync. Also preserve URLs, numbers, technical terms. Do not add any extra commentary or labels. The text is a description of an agent skill. Translate only the text inside <text>. The name "lipsync" appears in the first sentence as "Lip-sync" - we should preserve it as "lipsync" (or "Lip-sync"? The instruction says "Name to preserve: lipsync" so we should keep it as "lipsync" but the source has "Lip-sync" with hyphen and capital L. Probably we should keep the exact form from source? The instruction says "Preserve product names, protocol names, URLs, numbers, and technical terms." So we preserve "Lip-sync" as is? But the name to preserve is "lipsync" (lowercase). However, in

We need to translate the given text into Korean, preserving the specified names: lipsync, RunComfy, runcomfy, ByteDance OmniHuman, Sync Labs sync v2 / Pro, Kling lipsync, Creatify lipsync. Also preserve URLs, numbers, technical terms. Do not add any extra commentary or labels. The text is a description of an agent skill. Translate only the text inside <text>. The name "lipsync" appears in the first line as "Lip-sync" but we preserve it as "lipsync" (the name to preserve). Actually the instruction says "Name to preserve: lipsync" so we should keep that exact casing? The source has "Lip-sync" but we should preserve the name as given? The instruction says "preserve product names, protocol names, URLs, numbers, and technical terms." And "Name to preserve: lipsync" - likely we should keep it as "lipsync" even if source has hyphen? But the source says "Lip-sync" - I think we should keep

Agent Media OS — resolve any media need (BGM, SFX, image, icon) into a frozen local file + ledger record. One verb (`resolve`) handles the full cascade — project cache, global cache, HeyGen catalog search, freeze, register. Keeps search noise on disk, hands the agent a path. Use when a composition needs background music, sound effects, images, or icons.

Mediabunny 라이브러리를 사용한 멀티미디어 처리

Google Nano Banana 2(Gemini 계열 플래시 등급 텍스트-투-이미지)로 RunComfy에서 이미지를 생성합니다. 이 스킬은 모델의 문서화된 프롬프트 패턴과 함께 제공되어, 동일한 모델에 단순 프롬프트를 사용하는 것보다 더 정교한 출력을 얻을 수 있습니다. Nano Banana 2의 강점(빠른 반복, 이미지 내 타이포그래피 렌더링, 예측 가능한 프레임 구성, 선택적 웹 기반 컨텍스트), 해상도 등급별 가격, 안전 허용 범위 조절, 그리고 Nano Banana Pro / GPT Image 2 / Flux 2 / Seedream으로 전환해야 하는 경우를 설명합니다.

inference.sh CLI를 통해 Pruna P-Image 모델로 이미지를 생성합니다. 모델: P-Image, P-Image-LoRA, P-Image-Edit, P-Image-Edit-LoRA. 기능: 텍스트-이미지, 이미지 편집, LoRA 스타일, 다중 이미지 합성, 빠른 추론. Pruna는 품질 저하 없이 모델 속도를 최적화합니다. 트리거: pruna, p-image, pruna image, fast image generation, optimized flux, pruna ai, p image, fast ai image, economic image generation, cheap image generation

inference.sh CLI를 통해 Pruna P-Video 및 WAN 모델로 비디오를 생성합니다. 모델: P-Video, WAN-T2V, WAN-I2V. 기능: 텍스트-비디오, 이미지-비디오, 오디오 지원, 720p/1080p, 빠른 추론. Pruna는 품질 저하 없이 모델 속도를 최적화합니다. 트리거: pruna video, p-video, pruna ai video, fast video generation, optimized video, wan t2v, wan i2v, economic video generation, cheap video generation, pruna text to video, pruna image to video

inference.sh CLI를 통해 Pruna P-Video 및 WAN 모델로 비디오를 생성합니다. 모델: P-Video, WAN-T2V, WAN-I2V. 기능: 텍스트-비디오, 이미지-비디오, 오디오 지원, 720p/1080p, 빠른 추론. Pruna는 품질 저하 없이 모델 속도를 최적화합니다. 트리거: pruna video, p-video, pruna ai video, fast video generation, optimized video, wan t2v, wan i2v, economic video generation, cheap video generation, pruna text to video, pruna image to video

Pruna P-Video-Avatar를 inference.sh CLI를 통해 사용하여 말하는 헤드 아바타 동영상을 생성합니다. 초상화 이미지를 내장 TTS로 사실적인 말하는 동영상으로 변환합니다. 경쟁사 대비 18배 빠르고 6배 저렴합니다. 모델: P-Video-Avatar, P-Image(초상화 생성용). 기능: 텍스트-아바타, 오디오 기반 아바타, 30개 음성, 10개 언어, 720p/1080p, 내장 TTS, 동적 배경, 전신 제어. 용도: AI 프레젠터, 제품 데모, 설명 동영상, 가상 인플루언서, 마케팅 등.

Pruna P-Video-Avatar를 inference.sh CLI를 통해 사용하여 말하는 헤드 아바타 동영상을 생성합니다. 초상화 이미지를 내장 TTS로 사실적인 말하는 동영상으로 변환합니다. 경쟁사 대비 18배 빠르고 6배 저렴합니다. 모델: P-Video-Avatar, P-Image(초상화 생성용). 기능: 텍스트-아바타, 오디오 기반 아바타, 30개 음성, 10개 언어, 720p/1080p, 내장 TTS, 동적 배경, 전신 제어. 용도: AI 프레젠터, 제품 데모, 설명 동영상, 가상 인플루언서, 마케팅 등.

AI 비디오 생성 스킬로, Seedance 2, Kling 3.0, HappyHorse 등 10개 이상의 모델을 자동 선택합니다. 텍스트, 이미지, URL, 스크립트 또는 오디오( AI 음악, 립싱크, 멀티샷 시퀀싱 포함)로부터 완성된 멀티샷 비디오(5~120초)를 제작합니다. 프롬프트 작성이나 모델 선택이 필요 없습니다. 사용 용도: 비디오 제작, AI 비디오, 비디오 만들기, 제품 비디오, 브랜드 비디오, 프로모션 클립, 설명 비디오, 짧은 비디오, TikTok 비디오, Instagram Reel, YouTube Short, 제품 광고 등.

Relight a still image — change the lighting setup, color temperature, direction, or mood — on RunComfy via the `runcomfy` CLI. Routes to Qwen Edit 2509's dedicated `relight` LoRA endpoint for purpose-built relighting, with fallback to identity-preserving edit endpoints (Nano Banana 2 Edit, GPT Image 2 Edit, FLUX Kontext Pro) when prose lighting language is enough. Use for product relighting (studio softbox → window light), portrait mood shift (overcast → golden hour), or color-grade change....

Relight a still image — change the lighting setup, color temperature, direction, or mood — on RunComfy via the `runcomfy` CLI. Routes to Qwen Edit 2509's dedicated `relight` LoRA endpoint for purpose-built relighting, with fallback to identity-preserving edit endpoints (Nano Banana 2 Edit, GPT Image 2 Edit, FLUX Kontext Pro) when prose lighting language is enough. Use for product relighting (studio softbox → window light), portrait mood shift (overcast → golden hour), or color-grade change....

Relight a still image — change the lighting setup, color temperature, direction, or mood — on RunComfy via the `runcomfy` CLI. Routes to Qwen Edit 2509's dedicated `relight` LoRA endpoint for purpose-built relighting, with fallback to identity-preserving edit endpoints (Nano Banana 2 Edit, GPT Image 2 Edit, FLUX Kontext Pro) when prose lighting language is enough. Use for product relighting (studio softbox → window light), portrait mood shift (overcast → golden hour), or color-grade change....

Remotion에서 캡션 처리하기