DD
I started a challenge called Daily Drops β building small, functional tools every day using vibe coding. This DD challenge helps me: Visualize progress, Force the 80% rule, Learn by building
Here's what I've built so far.
DD1: ACT Matrix Image Maker
What it does: A tool that generates shareable images of the ACT (Acceptance & Commitment Therapy) Matrix β a psychology framework with 4 quadrants for self-reflection.
What I learned:
- PWA basics: manifest.json, service workers, app icons
- Deploying static sites to Netlify via drag-and-drop
DD2: γγγγγγ₯γΌγΉγ‘γΌγ«γΌ (Easy News Maker)
What it does: Paste any article (Korean, English, or Japanese) and get a simplified Japanese summary at JLPT N4-N3 level β complete with furigana readings and text-to-speech.
What I learned:
- Integrating Claude API in browser-based apps
- API key management: usage limits, spending alerts, localStorage storage
- HTML/CSS/JavaScript concepts
- Claude API (Sonnet) for summarization and simplification
DD3: Timestamp
What it does: Adds polaroid-style timestamps to photos. Extracts date/time from EXIF metadata automatically, manual input fallback. Square output with center-crop.
What I learned: Canvas image manipulation, EXIF.js for metadata extraction, iOS Safari native "save to photos" eliminates complex download workflows.
DD4: worksdrop
What it does: Minimal cross-device todo list. Eisenhower matrix tagging (1-4), optional deadlines, Google Sheets backend for sync.
What I learned: Google Apps Script as serverless API - simpler than direct Sheets API (no OAuth setup). Trade-off: slow response times.
DD5: FRUITS CHECKER
What it does: Inventory checker for fruit store. Upload delivery list images, Claude API extracts items via OCR
What I learned: Claude API image-to-text extraction. Environment auto-detection utility (window.storage vs localStorage). UX for physical contexts.
DD6: relationship drawing
What it does: Research for drama competition - character relationship diagram tools.
What I learned: Research-only DD is valid. Not everything needs implementation.
DD7: retrospective prototype writer
What it does: Auto-generates weekly retrospective drafts from daily sources (blog, AI School wrap-ups, Garmin data). Design phase only.
What I learned: Two-model approach: HuggingFace for sentiment scoring, Claude API for pattern interpretation. Separation of measurement vs insight.