rikkahub

RikkaHub is an Android native LLM chat client that lets you switch between multiple AI providers (OpenAI-compatible, Google, Anthropic-compatible APIs, and more) within a single app. It supports multimodal input, memory features, prompt variables, and per-provider customization so you can tailor conversations to different models or vendors. This MCP instance represents the client-side application rather than a traditional server; there isn’t a separate server process to run. Instead, you install and run the Android app on an Android device or emulator, and connect to provider APIs via your configured keys and endpoints.

To use the tools and capabilities, install the app on your device or emulator, then configure provider accounts under the settings panel. You can add custom API URLs, API keys, and model names to switch between providers, enable web access, enable memory features, and use the memory, translation, and prompt-variables features during conversations. The app also supports multimodal inputs (images, documentation, PDFs) and a variety of UI and accessibility options for seamless chatting across providers.

Prerequisites:

• Java Development Kit (JDK 11 or newer)
• Android Studio (recommended) or command-line build tools
• An Android device or emulator for testing
• Access to provider APIs (API keys/endpoints) you plan to use with the app

Step-by-step installation:

1. Clone the repository git clone https://github.com/your-org/rikkahub.git cd rikkahub

2. Install Android Studio and open the project

   • Launch Android Studio
   • Choose Open an existing Android Studio project and select the rikkahub directory
   • Wait for Gradle sync and dependencies to resolve

3. Obtain and place required configuration files

   • If the app requires google-services.json for Firebase/Analytics integration, place it under app/ as described in the repository notes
   • Ensure you have provider API keys configured in the app’s settings or in a local config file as per the project guidelines

4. Build and run

   • In Android Studio, click Run (or press Shift+F10) to build and deploy to a connected device or emulator
   • Alternatively, from the command line: ./gradlew assembleDebug adb install -r app/build/outputs/apk/debug/app-debug.apk

5. Configure providers

   • Launch the app on the device/emulator
   • Go to Settings > Providers and add your API keys/endpoints, model names, and any custom headers or bodies as needed

6. Test a session

   • Start a new chat, choose a provider from the models picker, and begin a conversation
   • Explore features like memory, translation, prompt variables, and multimodal input within chats

Notes and tips:

• The project is an Android client; there is no standalone MCP server component to run server-side. Provider connectivity is handled through API keys and configured endpoints within the app.
• If you encounter build or runtime issues, ensure Google Services configuration (if used) is correct and that all provider endpoints are reachable from your test environment.
• Place google-services.json in the app folder if your build requires Firebase/Analytics integration; otherwise this file is optional.
• For debugging provider issues, enable verbose logging in the app settings to capture network requests and responses for diagnosis.
• Keep API keys secure; do not commit them to version control. Use local configuration or environment-secured storage as supported by the project.