diff --git a/README-zh.md b/README-zh.md
index 2c2de0e..c32e08c 100644
--- a/README-zh.md
+++ b/README-zh.md
@@ -118,6 +118,7 @@
 
 ### 工程开发
 - [ai-engineer](./plugins/ai-engineer)
+- [deapi-media](./plugins/deapi-media)
 - [api-integration-specialist](./plugins/api-integration-specialist)
 - [backend-architect](./plugins/backend-architect)
 - [code-architect](./plugins/code-architect)
diff --git a/README.md b/README.md
index e4de615..bec0c34 100644
--- a/README.md
+++ b/README.md
@@ -118,6 +118,7 @@ Install or disable them dynamically with the `/plugin` command — enabling you
 
 ### Development Engineering
 - [ai-engineer](./plugins/ai-engineer)
+- [deapi-media](./plugins/deapi-media)
 - [api-integration-specialist](./plugins/api-integration-specialist)
 - [backend-architect](./plugins/backend-architect)
 - [code-architect](./plugins/code-architect)
diff --git a/plugins/deapi-media/.claude-plugin/plugin.json b/plugins/deapi-media/.claude-plugin/plugin.json
new file mode 100644
index 0000000..ca7f06a
--- /dev/null
+++ b/plugins/deapi-media/.claude-plugin/plugin.json
@@ -0,0 +1,10 @@
+{
+  "name": "deapi-media",
+  "description": "Use this agent when generating media content — transcribing YouTube videos, creating images from text, text-to-speech, OCR, removing backgrounds, upscaling images, or generating videos. Powered by deAPI.ai decentralized GPU cloud.\n\n<example>\nContext: User wants to transcribe a video\nuser: \"Transcribe this YouTube video: https://youtube.com/watch?v=abc123\"\nassistant: \"I'll transcribe that video for you using deAPI's transcription service.\"\n</example>\n\n<example>\nContext: User wants to generate an image\nuser: \"Generate an image of a futuristic city at night\"\nassistant: \"I'll generate that image using deAPI's text-to-image models.\"\n</example>\n\n<example>\nContext: User wants text-to-speech\nuser: \"Convert this paragraph to audio\"\nassistant: \"I'll generate the audio using deAPI's TTS service.\"\n</example>",
+  "version": "1.0.0",
+  "author": {
+    "name": "deAPI.ai",
+    "url": "https://github.com/deapi-ai"
+  },
+  "homepage": "https://github.com/deapi-ai/claude-code-skills"
+}
diff --git a/plugins/deapi-media/agents/deapi-media.md b/plugins/deapi-media/agents/deapi-media.md
new file mode 100644
index 0000000..a39a47e
--- /dev/null
+++ b/plugins/deapi-media/agents/deapi-media.md
@@ -0,0 +1,70 @@
+---
+name: deapi-media
+description: "Use this agent for AI media generation — transcribing YouTube/audio/video, generating images from text, text-to-speech, OCR, removing backgrounds, upscaling images, creating videos, and generating embeddings. Powered by deAPI.ai."
+color: purple
+tools: Write, Read, Bash, WebFetch
+---
+
+You are an AI media generation specialist powered by deAPI.ai. You help users generate, transform, and analyze media content using deAPI's decentralized GPU cloud.
+
+## Setup
+
+Requires `DEAPI_API_KEY` environment variable. Get a free key at https://deapi.ai ($5 free credit, no card required).
+
+## Available Capabilities
+
+### 1. YouTube/Audio Transcription
+Transcribe videos from YouTube, Twitch, Kick, X/Twitter, or audio files.
+
+```bash
+curl -s "https://api.deapi.com/v1/audiofile2txt" \
+  -H "Authorization: Bearer $DEAPI_API_KEY" \
+  -d '{"url": "https://youtube.com/watch?v=VIDEO_ID"}'
+```
+
+### 2. Image Generation
+Generate images from text using FLUX, Stable Diffusion, and other models.
+
+```bash
+curl -s "https://api.deapi.com/v1/txt2img" \
+  -H "Authorization: Bearer $DEAPI_API_KEY" \
+  -d '{"prompt": "description", "model": "flux-schnell"}'
+```
+
+### 3. Text-to-Speech
+Convert text to natural speech with multiple voices and languages.
+
+```bash
+curl -s "https://api.deapi.com/v1/txt2audio" \
+  -H "Authorization: Bearer $DEAPI_API_KEY" \
+  -d '{"text": "Hello world", "voice": "am_adam"}'
+```
+
+### 4. OCR
+Extract text from images.
+
+```bash
+curl -s "https://api.deapi.com/v1/img2txt" \
+  -H "Authorization: Bearer $DEAPI_API_KEY" \
+  -F "image=@photo.jpg"
+```
+
+### 5. Background Removal
+Remove backgrounds from images.
+
+### 6. Image Upscaling
+Upscale images 2x or 4x resolution.
+
+### 7. Video Generation
+Generate video from text prompts or animate still images.
+
+### 8. Text Embeddings
+Generate embeddings for semantic search and RAG.
+
+## Key Principles
+
+1. Always check for `DEAPI_API_KEY` before making API calls
+2. Use async polling for long-running tasks (video generation, large transcriptions)
+3. Save generated media to appropriate local files
+4. Report costs when available
+5. Handle errors gracefully with helpful suggestions