From Overwhelmed Reader to Automated Sharer: How I Built a macOS App to Bridge Reading and Posting
Published on August 20, 2026
After accumulating over 200 unread articles in my "Later" list, I realized that saving content had become an overwhelming deposit rather than a productive habit. To solve the friction between reading and sharing, I built a native macOS application called Reader Chomper that integrates Readwise Reader, large language models, and social media APIs into a single workflow.
Key Takeaways
- The Problem of Volume: Accumulating hundreds of unread articles creates decision paralysis, leading to saved content that is never consumed or shared.
- Triage via Limits: Limiting the view to 10 articles at a time forces finite decision-making and prevents overwhelm compared to viewing an entire library.
- Quick Scan Efficiency: A 100-word summary with three bullet points allows for rapid assessment, often eliminating the need to read the full article.
- Contextual Depth: Integrating an LLM allows users to chat directly with the source material to determine relevance without leaving the app.
- Frictionless Sharing: Automating the draft generation and queuing process via API removes the barrier between reading and publishing on platforms like LinkedIn and Bluesky.
The Workflow: Four Steps to Publication
Reader Chomper operates through four distinct steps: triage, quick scan, deep summary, and sharing. This sequence helps decide what is worth time and ensures posts are queued before moving to the next article.
Step 1: Triage 10 at a Time
When I open Reader Chomper, it pulls exactly 10 articles from the top of my "Later" list in Readwise Reader and stops there. If I need more content, I must manually hit "Refill." This step was the most challenging to engineer because Readwise’s default view displays everything at once, which previously caused me to feel overwhelmed and close the tab. By limiting the view to 10 items, the session feels finite. I view this not as a restriction but as a contract with myself: read 10 articles, decide what to do with each, and consider the session complete for the day.
Step 2: Run a Quick Scan
Upon tapping into an article, Reader Chomper generates a "Quick Scan": a 100-word summary containing three bullet points with zero fluff. For most articles, this is where my engagement ends. Once I have the three main points, I can decide to move on. This step has made my reading habits more honest; I no longer convince myself I will "come back" to articles I know I won't read. The decision now takes about 30 seconds: either I have what I need and archive the article, or there is something worth pursuing.
Step 3: Deep Summary and Chat
If the three bullets from the Quick Scan are insufficient, I have two options for deeper engagement. The first is a longer summary of approximately 300 words. The second, which I use most often, is chatting with the article via Claude. For example, when I encountered an article titled "DuckLake for busy engineering managers," the Quick Scan explained what DuckLake was, but not its specific relevance to my role. I asked Claude, "Why is DuckLake relevant to engineering managers?" and received a contextual answer pulled directly from the source in two seconds. This feature allows me to ask specific questions that many read-it-later apps do not support.
Step 4: Queue to
When an article warrants a post, I hit "Share." Reader Chomper generates a draft consisting of a short blurb capturing the key idea and the link. If the tone isn't quite right, I can hit "Regenerate" or edit it manually. The API handles the posting instantly, archiving the article in Readwise while the post sits in the queue.
The Technical Stack
The architecture behind Reader Chomper is intentionally simple, relying on tools I was already paying for or using.
- SwiftUI and Native macOS: I considered a browser extension or web app for faster development, but I prioritized performance and aesthetics. A native app launches instantly, scrolls without lag, and integrates seamlessly with the macOS interface.
- Readwise Reader: This remains the source of truth for my saved articles. Reader Chomper acts as a better interface for processing this queue; any article archived in Reader Chomper is also archived in Readwise, keeping both systems clean.
- LLMs (Anthropic and OpenAI): I use both APIs interchangeably to summarize articles and generate blurbs. The app displays the running cost of these API calls in the corner, a small detail that reminds me that these requests are not free and helps identify when I am bouncing between articles unnecessarily.
- SQLite and Dropbox: Article data, summaries, and chat history are stored in an SQLite file within my Dropbox folder. This ensures the database syncs across devices without requiring custom synchronization infrastructure.
Results and Impact
My posting frequency has not increased; I still aim for two posts a week on LinkedIn and Bluesky. However, the quality of what I share has improved significantly. Previously, my followers received bare links that required them to click through to understand the value. Now, they receive a take that does some of that work for them. The AI summary surfaces an idea, allowing me to form my own perspective based on that context.
The queue often runs months deep, allowing me to build up posts during heavy reading weeks and coast during lighter ones. The gap between reading something good and posting about it is now closed. I read more thoughtfully because I know I have a place to put my opinions, confirming that this automation is doing its job effectively.