Generate marketing ad banners with LINE, Gemini, and Nano Banana Pro
About This Template This workflow creates high-quality, text-rich advertising banners from simple LINE messages. It combines Google Gemini (for marketing-focused prompt engineering) and Nano Banana Pro (accessed via Kie.ai API) to generate images with superior text rendering capabilities. It also handles the asynchronous API polling required for high-quality image generation. How It Works Input: Users send a banner concept via LINE (e.g., "Coffee brand, morning vibe"). Prompt Engineering: Gemini optimizes the request into a detailed prompt, specifying lighting, composition, and Japanese catch-copy placement. Async Generation: The workflow submits a job to Nano Banana Pro (Kie API) and intelligently waits/polls until the image is ready. Hosting: The final image is downloaded and uploaded to a public AWS S3 bucket. Delivery: The image is pushed back to the user on LINE. Who It’s For Marketing teams creating A/B test assets. Japanese market advertisers needing accurate text rendering. Developers looking for an example of Async API Polling patterns in n8n. Requirements n8n (Cloud or Self-hosted). Kie.ai API Key (for Nano Banana Pro model). Google Gemini API Key. AWS S3 Bucket (Public access enabled). LINE Official Account (Messaging API). Setup Steps Credentials: Configure the "Header Auth" credential for the Kie.ai nodes (Header: Authorization, Value: Bearer YOURAPIKEY). AWS: Ensure your S3 bucket allows public read access so LINE can display the image. Webhook: Add the production webhook URL to your LINE Developers console.
By Masaki GoCreate Jira tickets from Streamlit forms with webhook & REST API
Description Automated workflow that creates Jira issues directly from Streamlit form submissions. Receives webhook data, validates and transforms it to Jira's API schema, creates the issue, and returns the ticket details to the frontend application. Context Bridges the gap between lightweight Streamlit prototypes and enterprise Jira workflows. Enables rapid ticket creation while maintaining Jira as the authoritative source of truth. Includes safety mechanisms to prevent duplicate submissions and malformed requests. Target Users Product Managers building internal request portals. Engineering Managers creating demo applications. Teams requiring instant Jira integration without complex UI development. Project Manager using Jira pour mangement and reporting. Organizations wanting controlled ticket creation without exposing Jira directly. Technical Requirements n8n instance (cloud or self-hosted) with webhook capabilities Jira Cloud project with API token and issue creation permissions Streamlit application configured to POST to n8n webhook endpoint Optional: Custom field IDs for Story Points (typically customfield_10016) Workflow Steps Webhook Trigger - Receives POST from Streamlit with ticket payload. Deduplication Guard - Filters out ping requests and rapid duplicate submissions. Data Validation - Ensures required fields are present and properly formatted. Schema Transformation - Maps Streamlit fields to Jira API structure. Jira API Call - Creates issue via REST API with error handling. Response Formation - Returns success status with issue key and URL. Key Features Duplicate submission prevention. Rich text description formatting for Jira. Configurable priority and issue type mapping. Story points integration for agile workflows. Comprehensive error handling and logging. Clean JSON response for frontend feedback. Validation Testing Ping/test requests are ignored without creating issues. First submission creates Jira ticket with proper formatting. Rapid resubmission is blocked to prevent duplicates. All field types (priority, labels, due dates, story points) map correctly. Error responses are handled gracefully. Expected Output Valid Jira issue created in specified project JSON response: {ok: true, jiraKey: "PROJ-123", url: "https://domain.atlassian.net/browse/PROJ-123"} No orphaned or duplicate tickets. Audit trail in n8n execution logs. Implementation Notes Jira Cloud requires accountId for assignee (not username). Date format must be YYYY-MM-DD for due dates. Story Points field ID varies by Jira configuration. Enable response output in HTTP node for debugging. Consider rate limiting for high-volume scenarios. Tutorial video: Watch the Youtube Tutorial video How it works ⏰ Trigger: Webhook fires when the app submits. 🧹 Guard: Ignore pings/invalid, deduplicate rapid repeats. 🧱 Prepare: Normalize to Jira’s field model (incl. Atlassian doc description). 🧾 Create: POST to /rest/api/3/issue and capture the key. 🔁 Respond: Send { ok, jiraKey, url } back to Streamlit for instant UI feedback. About me : I’m Yassin a Project & Product Manager Scaling tech products with data-driven project management. 📬 Feel free to connect with me on Linkedin
By Yassin Zehar