Gemini AI

Send prompts to Google Gemini from Homey Flows and use the AI's answer in your automations. Features - Text Prompts: "Send Prompt" action card that accepts text prompts and returns AI-generated responses; - Image Analysis: "Send Prompt with Image" action card that accepts multimodal prompts (image + text) and returns AI-generated responses; - Simple settings page to store your Gemini API Key; - Powered by Google Generative AI (gemini-2.5-flash-lite). Usage Examples ``` WHEN: Motion is detected in the living room THEN: Send Prompt "Generate a welcoming message for someone entering the living room" AND: Speak the Gemini answer ``` ``` WHEN: Weather changes THEN: Send Prompt "Create a brief weather advisory based on today's forecast" AND: Send notification with Gemini answer ``` ``` WHEN: Doorbell camera detects motion THEN: Send Prompt with Image "Describe what you see in this image and identify any people or packages" AND: Send notification with Gemini analysis ``` ``` WHEN: Security sensor triggers THEN: Take a snapshot with camera AND: Send Prompt with Image "Analyze this security camera image and describe any potential threats" AND: Log the analysis result ``` ``` WHEN: Motion detected in storage room THEN: Capture image from security camera AND: Send Prompt with Image "Check if anything appears disturbed or out of place in this storage area. Reply with 'true' if everything is OK, otherwise reply with 'false'" AND: Manage true/false response ``` Requirements - A valid Google Gemini API Key (see setup guide: https://github.com/s-dimaio/com.dimapp.geminiforhomey#getting-your-google-gemini-api-key). Setup 1) Open the app’s Settings and insert your API Key. 2) Create a Flow and add the action “Send Prompt” or “Send Prompt with Image”. 3) Provide the prompt and use the “Gemini answer” token in subsequent Flow cards. Privacy - The app stores only your API Key in Homey’s settings. - Prompts and answers are sent to Google’s API when you run the Flow.