Add new tools: generate image, edit image, remove background, upscale image

main.py

@@ -1,6 +1,6 @@
 from dify_plugin import Plugin, DifyPluginEnv
 
-plugin = Plugin(DifyPluginEnv(MAX_REQUEST_TIMEOUT=120))
+plugin = Plugin(DifyPluginEnv(MAX_REQUEST_TIMEOUT=600))
 
 if __name__ == '__main__':
     plugin.run()

manifest.yaml

@@ -10,7 +10,7 @@ label:
   en_US: MuAPI
 
 description:
-  en_US: Generate images and videos using MuAPI
+  en_US: Generate images, videos, and audio using MuAPI. Edit images, remove backgrounds, upscale images, and remix audio with 100+ AI models.
 
 created_at: "2026-06-12T00:00:00.000000Z"
 

provider/muapi.yaml

@@ -2,7 +2,7 @@ identity:
   author: muapi
   name: muapi
   description:
-    en_US: Generate images and videos using MuAPI
+    en_US: Generate images, videos, and audio using MuAPI. Edit images, remove backgrounds, upscale, and remix audio.
   icon: icon.png
   label:
     en_US: MuAPI
@@ -16,6 +16,13 @@ credentials_for_provider:
 
 tools:
   - tools/generate_image.yaml
+  - tools/generate_video.yaml
+  - tools/image_to_video.yaml
+  - tools/edit_image.yaml
+  - tools/remove_background.yaml
+  - tools/upscale_image.yaml
+  - tools/generate_audio.yaml
+  - tools/remix_audio.yaml
 
 extra:
   python:

provider/muapi_provider.py

@@ -1,6 +1,10 @@
+import os
+
 from dify_plugin import ToolProvider
 from dify_plugin.errors.tool import ToolProviderCredentialValidationError
 
+from muapi import MuAPI
+
 
 class MuapiProvider(ToolProvider):
 
@@ -19,7 +23,13 @@ def validate_credentials(
                     "MuAPI API key is required"
                 )
 
+            os.environ["MUAPI_API_KEY"] = api_key
+
+            client = MuAPI()
+
+            client.accounts.balance()
+
         except Exception as e:
             raise ToolProviderCredentialValidationError(
-                str(e)
+                f"MuAPI credential validation failed: {str(e)}"
             )

requirements.txt

@@ -1,2 +1,2 @@
-dify-plugin
+dify-plugin==0.9.1
 git+https://github.com/SamurAIGPT/muapi-python.git

tools/edit_image.py

@@ -0,0 +1,55 @@
+from collections.abc import Generator
+from typing import Any
+import os
+
+from dify_plugin import Tool
+from dify_plugin.entities.tool import ToolInvokeMessage
+
+from muapi import MuAPI
+
+
+class EditImageTool(Tool):
+    def _invoke(
+        self,
+        tool_parameters: dict[str, Any],
+    ) -> Generator[ToolInvokeMessage]:
+
+        api_key = self.runtime.credentials.get(
+            "muapi_api_key"
+        )
+
+        if not api_key:
+            yield self.create_json_message(
+                {
+                    "error": "Missing MuAPI API key"
+                }
+            )
+            return
+
+        os.environ["MUAPI_API_KEY"] = api_key
+
+        client = MuAPI()
+
+        kwargs = {
+            "image": tool_parameters["image"],
+            "prompt": tool_parameters["prompt"],
+            "model": tool_parameters.get(
+                "model",
+                "flux-kontext-dev"
+            ),
+        }
+
+        aspect_ratio = tool_parameters.get("aspect_ratio")
+        if aspect_ratio:
+            kwargs["aspect_ratio"] = aspect_ratio
+
+        result = client.images.edit(**kwargs)
+
+        if isinstance(result, dict) and result.get("images"):
+            yield self.create_image_message(
+                result["images"][0]
+            )
+
+        yield self.create_json_message(
+            result if isinstance(result, dict) else {"result": str(result)}
+        )

tools/edit_image.yaml

@@ -0,0 +1,56 @@
+identity:
+  name: edit_image
+  author: muapi
+  label:
+    en_US: Edit Image
+
+description:
+  human:
+    en_US: Edit an existing image using a text prompt via MuAPI
+  llm: Edit or transform an existing image based on a text prompt using AI image editing models via MuAPI
+
+parameters:
+  - name: image
+    type: string
+    required: true
+    form: llm
+    label:
+      en_US: Image URL
+    human_description:
+      en_US: URL of the image to edit
+    llm_description: URL of the source image to edit or transform
+
+  - name: prompt
+    type: string
+    required: true
+    form: llm
+    label:
+      en_US: Prompt
+    human_description:
+      en_US: Text prompt describing the desired edit
+    llm_description: Text prompt describing how the image should be edited or transformed
+
+  - name: model
+    type: string
+    required: false
+    form: form
+    label:
+      en_US: Model
+    human_description:
+      en_US: The model to use for image editing (e.g. flux-kontext-dev)
+    llm_description: The model to use for image editing
+    default: flux-kontext-dev
+
+  - name: aspect_ratio
+    type: string
+    required: false
+    form: form
+    label:
+      en_US: Aspect Ratio
+    human_description:
+      en_US: Aspect ratio of the output image (e.g. 1:1, 16:9, 4:3)
+    llm_description: Aspect ratio of the edited output image
+
+extra:
+  python:
+    source: tools/edit_image.py

tools/generate_audio.py

@@ -0,0 +1,64 @@
+from collections.abc import Generator
+from typing import Any
+import os
+
+from dify_plugin import Tool
+from dify_plugin.entities.tool import ToolInvokeMessage
+
+from muapi import MuAPI
+
+
+class GenerateAudioTool(Tool):
+    def _invoke(
+        self,
+        tool_parameters: dict[str, Any],
+    ) -> Generator[ToolInvokeMessage]:
+
+        api_key = self.runtime.credentials.get(
+            "muapi_api_key"
+        )
+
+        if not api_key:
+            yield self.create_json_message(
+                {
+                    "error": "Missing MuAPI API key"
+                }
+            )
+            return
+
+        os.environ["MUAPI_API_KEY"] = api_key
+
+        client = MuAPI()
+
+        result = client.audio.create(
+            prompt=tool_parameters["prompt"],
+            title=tool_parameters.get(
+                "title",
+                ""
+            ),
+            tags=tool_parameters.get(
+                "tags",
+                ""
+            ),
+            instrumental=bool(
+                tool_parameters.get(
+                    "instrumental",
+                    False
+                )
+            ),
+        )
+
+        audio_url = None
+
+        if isinstance(result, dict):
+            audio_url = result.get("audio") or result.get("url") or result.get("output")
+
+            if not audio_url and result.get("audio_url"):
+                audio_url = result["audio_url"]
+
+        if audio_url:
+            yield self.create_link_message(audio_url)
+
+        yield self.create_json_message(
+            result if isinstance(result, dict) else {"result": str(result)}
+        )

tools/generate_audio.yaml

@@ -0,0 +1,56 @@
+identity:
+  name: generate_audio
+  author: muapi
+  label:
+    en_US: Generate Audio
+
+description:
+  human:
+    en_US: Generate audio or music from a text prompt using MuAPI
+  llm: Generate audio or music from a text description using AI audio generation via MuAPI
+
+parameters:
+  - name: prompt
+    type: string
+    required: true
+    form: llm
+    label:
+      en_US: Prompt
+    human_description:
+      en_US: Text prompt describing the audio or music to generate
+    llm_description: Text prompt describing the audio or music to generate
+
+  - name: title
+    type: string
+    required: false
+    form: form
+    label:
+      en_US: Title
+    human_description:
+      en_US: Title for the generated audio track
+    llm_description: Title for the generated audio track
+
+  - name: tags
+    type: string
+    required: false
+    form: form
+    label:
+      en_US: Tags
+    human_description:
+      en_US: Comma-separated genre or style tags (e.g. pop, electronic, chill)
+    llm_description: Comma-separated genre or style tags for the audio
+
+  - name: instrumental
+    type: boolean
+    required: false
+    form: form
+    label:
+      en_US: Instrumental
+    human_description:
+      en_US: Generate instrumental only (no vocals)
+    llm_description: Whether to generate instrumental audio without vocals
+    default: false
+
+extra:
+  python:
+    source: tools/generate_audio.py

tools/generate_video.py

@@ -0,0 +1,64 @@
+from collections.abc import Generator
+from typing import Any
+import os
+
+from dify_plugin import Tool
+from dify_plugin.entities.tool import ToolInvokeMessage
+
+from muapi import MuAPI
+
+
+class GenerateVideoTool(Tool):
+    def _invoke(
+        self,
+        tool_parameters: dict[str, Any],
+    ) -> Generator[ToolInvokeMessage]:
+
+        api_key = self.runtime.credentials.get(
+            "muapi_api_key"
+        )
+
+        if not api_key:
+            yield self.create_json_message(
+                {
+                    "error": "Missing MuAPI API key"
+                }
+            )
+            return
+
+        os.environ["MUAPI_API_KEY"] = api_key
+
+        client = MuAPI()
+
+        result = client.videos.generate(
+            prompt=tool_parameters["prompt"],
+            model=tool_parameters.get(
+                "model",
+                "kling-master"
+            ),
+            duration=int(
+                tool_parameters.get(
+                    "duration",
+                    5
+                )
+            ),
+            aspect_ratio=tool_parameters.get(
+                "aspect_ratio",
+                "16:9"
+            ),
+        )
+
+        video_url = None
+
+        if isinstance(result, dict):
+            video_url = result.get("video") or result.get("url") or result.get("output")
+
+            if not video_url and result.get("videos"):
+                video_url = result["videos"][0]
+
+        if video_url:
+            yield self.create_link_message(video_url)
+
+        yield self.create_json_message(
+            result if isinstance(result, dict) else {"result": str(result)}
+        )