GenAI Tasker Plugin

Antivirus Warning

Some antivirus programs may flag this app as potentially harmful. This is a known issue that can occur with automation tools, plugins, or apps that use dynamic code execution or network APIs.

GenAI Tasker Plugin is fully open-source, and you are encouraged to review the code yourself.

There is no malicious functionality included in this project. However, if you are uncomfortable installing the app, you should trust your judgment and avoid doing so.

---

A versatile Android plugin for Tasker that allows you to integrate various Generative AI models into your automation workflows.

Features

• Multiple Providers: Support for OpenAI, Google Gemini, Anthropic Claude, OpenRouter, xAI (Grok), and local Ollama instances.
• Vision Support: Upload local images or use Tasker variables containing image URIs for multi-modal interactions.
• Customizable: Override default models and API base URLs (perfect for local proxies or self-hosted LLMs).
• Flexible Input:
  • Manual Mode: Build conversations directly in the Tasker UI.
  • Variable/JSON Mode: Pass dynamic message arrays or raw text using Tasker variables.
• Tasker Integration: Seamlessly returns AI responses to Tasker variables (default: %ai_response).

Supported Providers

Provider	Default Model	Custom Base URL Support
OpenAI	gpt-4o-mini	Yes
Gemini	gemini-1.5-flash	Yes
Claude	claude-3-5-sonnet-20240620	Yes
OpenRouter	google/gemini-flash-1.5	Yes
xAI (Grok)	grok-2-latest	Yes
Ollama	llama3	Yes (Required for local host)

Installation

1. Download and install the APK.
2. Open Tasker.
3. Create a new Action.
4. Go to Plugin -> GenAI Tasker Plugin -> AI Action.
5. Tap the Configuration (pencil) icon to set up your request.

Configuration

1. Basic Setup

• Provider: Select your preferred AI service.
• API Key: Enter your secret key (not required for Ollama if running locally).
• Model: (Optional) Specify a specific model ID (e.g., gpt-4o, claude-3-opus, grok-2-latest).
• Base URL: (Optional) Provide a custom endpoint (e.g., http://192.168.1.10:11434 for Ollama).

2. Images (Vision)

• Image URI: Provide a local file path (e.g., /sdcard/Pictures/photo.jpg) or a Tasker variable like %image_path. The plugin will Base64 encode the image and attach it to the prompt.

3. Messages

• Manual Messages: Add multiple roles (System, User, Assistant) to build a conversation history.
• Variable / JSON: Toggle to this tab to pass a raw string or a JSON array of messages (e.g., [{"role": "user", "content": "Hello!"}]).

Variables Returned

• %ai_response: The text content of the AI's reply.
• %ai_raw_json: (Optional) The full raw JSON response from the provider for advanced parsing.

Requirements

• Android 8.0+
• Tasker installed and configured.
• Active internet connection (unless using local Ollama).

Development

Built with Kotlin and uses the Tasker Plugin Library.

Build

./gradlew assembleDebug

## License

MIT License - feel free to use and modify for your own automation needs!