Daniel J Wilson

Daniel J Wilson Writing on research, technology, building, and everyday life. 2026-02-09T00:00:00.000Z https://danieljwilson.com/ Daniel J Wilson An Email-to-Input Pipeline with Claude Code https://danieljwilson.com/writing/post_email-to-input-pipeline/ 2026-02-09T00:00:00.000Z I consume a lot of podcasts, books, movies, and articles. Keeping track of them was always a friction point — enough friction that I mostly didn't bother, which meant losing references I wanted to revisit later. So I built a simple pipeline: email a URL to myself, and a Claude Code skill turns it into a structured note. <hr> <h4>How it works</h4> The system has two parts: a Python script that fetches URLs from Gmail, and a Claude Code skill that processes each one. 1. Send an email. I forward or compose an email to a dedicated Gmail +inputs address. The body contains the URL. That's it. 2. The email subject controls metadata. The subject line is a simple flag system: <ul> <li>Empty — not finished, not recommended</li> <li><code>F</code> — finished consuming it</li> <li><code>R</code> — recommended but not finished</li> <li><code>FR</code> — finished and recommended</li> </ul> If the subject is <code>F</code> or <code>FR</code>, the finish date is automatically set to when the email was sent. 3. Run <code>/input-batch</code>. A Claude Code slash command triggers the pipeline. It connects to Gmail via IMAP, pulls unread emails from the +inputs address, extracts URLs, and deletes the processed emails. 4. Auto-detect and fetch metadata. Each URL is categorized by domain — Apple Podcasts becomes a podcast entry, Goodreads becomes a book, IMDB becomes a movie, and so on. Claude then fetches the page and extracts structured metadata: title, creator, duration, release date, cover art, and whatever else is relevant to the category. 5. Generate a summary. Each input gets a 2-4 sentence summary with key concepts bolded. 6. Create a markdown file. The note is written to my Obsidian vault as a markdown file with structured frontmatter — ready to be rendered on my website or queried later. 7. Commit and push. The content repo is automatically committed and pushed to GitHub, which triggers a site rebuild. <hr> <h4>Handling NotebookLM</h4> One edge case: I use <a href="https://notebooklm.google.com/" target="_blank" rel="noopener noreferrer">NotebookLM</a> to generate podcast-style audio summaries of papers and essays. These URLs require Google auth, so Claude can't fetch them directly. The workaround: I attach a screenshot of the NotebookLM summary page to the email. Claude reads the image to extract the title, author, and publication date, then searches the web for any missing metadata. <hr> <h4>Why email?</h4> Email is the universal &quot;share to&quot; target. Every app on my phone can share a link via email. No special app needed, no API to configure, no bookmarklet to maintain. The subject line flag system (<code>F</code>, <code>R</code>, <code>FR</code>) adds just enough metadata without any real friction. The whole thing takes about 5 seconds per input on my end. Everything else is automated. <hr> <h4>The stack</h4> <ul> <li>Gmail — inbox as a queue (IMAP access via Python)</li> <li>Claude Code — slash command skill for orchestration, metadata extraction, and summary generation</li> <li>Obsidian — markdown files with YAML frontmatter</li> <li>GitHub — content repo that triggers site deploys</li> </ul> Constipated Altruism https://danieljwilson.com/writing/post_constipated-altruism/ 2026-02-07T00:00:00.000Z While reading a recent <a href="https://substack.com/@experimentalhistory" target="_blank" rel="noopener noreferrer">Adam Mastroianni</a> <a href="https://www.experimental-history.com/p/underrated-ways-to-change-the-world" target="_blank" rel="noopener noreferrer">post</a> I came across a term/concept that struck a chord: Constipated altruism. The phrase is used to describe what seems to be a common failure mode: <blockquote> You want to help. You care. But you don’t actually do anything. </blockquote> Not because you’re selfish. Because you’ve unconsciously decided that only huge or heroic actions “count.” So unless you: <ul> <li>donate a life-changing amount of money,</li> <li>switch careers to work full-time on a cause,</li> <li>or launch something world-saving,</li> </ul> …you do nothing. The altruistic impulse has nowhere to go. It just sits there. Hence: constipated. <hr> <h4>The trap</h4> The hidden assumption is that impact must be either: <ol> <li>High-status — become rich/powerful first, then fix things</li> <li>High-sacrifice — give up comfort, stability, or your career</li> </ol> Everything else feels trivial, performative, or not worth doing. But that framing quietly eliminates 95% of the actions that actually move the world. Most change comes from: <ul> <li>small, repeatable help</li> <li>local improvements</li> <li>medium-scale efforts</li> <li>connecting people</li> <li>building tools</li> <li>sharing knowledge</li> <li>removing friction</li> <li>being the “second-bravest person” who supports someone else’s initiative</li> </ul> In other words: boring, doable help. <hr> <h4>Why this idea matters</h4> This framing removes guilt and replaces it with agency. Instead of: <blockquote> “This problem is too big, so nothing I do matters.” </blockquote> It becomes: <blockquote> “What’s the smallest concrete thing I could do today that slightly improves this?” </blockquote> That question is almost always answerable. An altruism action &quot;laxative&quot; of sorts? And once you allow small actions to count, action becomes easy again. Momentum beats moral purity. <hr> <h4>A simple heuristic I’m keeping</h4> When I notice myself “caring but not acting,” I try: Lower the bar until action feels trivial. Then do that. Tiny counts. Because the alternative isn’t “bigger impact.” It’s usually nothing. <hr> <h4>Original essay</h4> <ul> <li>Adam Mastroianni — <a href="https://www.experimental-history.com/p/underrated-ways-to-change-the-world" target="_blank" rel="noopener noreferrer">Underrated ways to change the world</a></li> </ul> <h4>Related links</h4> <ul> <li><a href="https://www.experimental-history.com/p/underrated-ways-to-change-the-world/comments" target="_blank" rel="noopener noreferrer">Comments/discussion on the post</a></li> <li><a href="https://substack.com/@amrita112/note/c-79142642" target="_blank" rel="noopener noreferrer">Short Substack note quoting the idea</a></li> <li><a href="https://www.metafilter.com/212060/Underrated-ways-to-change-the-world-Part-1" target="_blank" rel="noopener noreferrer">MetaFilter discussion</a></li> <li><a href="https://www.linkedin.com/posts/shyju-varkey-98282712_sharingstories-storytelling-whatireadlastweek-activity-7274680600718954496-y9If" target="_blank" rel="noopener noreferrer">LinkedIn reference</a></li> <li><a href="https://www.thecoloradojournal.com/journal-xxv/" target="_blank" rel="noopener noreferrer">Reprint/context mention</a></li> </ul> <hr> LLM code de-cruftification https://danieljwilson.com/writing/tips -llm-code-cleanup/ 2026-02-05T00:00:00.000Z <h4>Reducing LLM “cruft” after large refactors</h4> One simple strategy that’s worked well for me is a reviewer → verifier two-pass workflow. <pre><code class="language-less">Codebase ↓ Pass A: Reviewer → produces checklist only ↓ Pass B: Verifier → applies fixes + runs checks ↓ Clean repo </code></pre> <blockquote class="callout callout-note" data-callout="note"> ExampleHere is an example of the prompt I am currently using (I will try to keep it updated). It is current as of 2026-02-05 14:25. </blockquote> <pre><code>Use a reviewer + verifier paradigm to now go over all the code in [path]. Pass A (reviewer): Audit the refactor for dead code, duplicate paths, inconsistent patterns, and doc drift. Produce a checklist of fixes. Pass B (verifier): Apply fixes, run lint/format/tests/build, and update impacted READMEs. Report deletions and results. Goals: - Remove all obsolete/duplicate code created by the refactor. - Reduce complexity and redundancy (no parallel implementations). - Ensure repo consistency (naming, structure, patterns). - Update all impacted docs/READMEs. Notes: - Prefer deletion over deprecation. If something is unused, remove it. - No new abstractions unless they reduce duplication across ≥2 call sites. - No TODOs unless they include owner + next action + condition. - If you touch a public API, update examples and changelog notes. </code></pre> <h4>Why this works</h4> LLMs tend to add layers and leave residue during refactors. Separating analysis from execution forces: <ul> <li>less speculative code</li> <li>fewer unnecessary abstractions</li> <li>more aggressive deletion</li> <li>tighter repo consistency</li> <li>docs that actually match reality</li> </ul> Net effect: smaller diffs, simpler code, and a repo that stays clean instead of slowly accreting AI sludge. Option (⌥) + Shift (⇧) + 2 = € https://danieljwilson.com/writing/tips-euro-keyboard-shortcut/ 2026-02-04T00:00:00.000Z Given that I live in France right now, I have to type the € sign on a semi-regular basis. If you’re using a U.S. keyboard layout on macOS, typing the euro symbol isn’t obvious — it’s not printed on any key (in initial days I found myself searching for the symbol on the web and then copy pasting the it). But the shortcut is pretty simple: <code>Option (⌥) + Shift (⇧) + 2</code> Press all three together → € <h4>Why this combo?</h4> On the U.S. layout: <ul> <li><code>Shift + 2</code> normally gives <code>@</code></li> <li>Holding <code>Option</code> switches the keyboard to the alternate symbol layer</li> <li>So <code>Option + Shift + 2</code> maps to €</li> </ul> <h4>Quick memory trick</h4> Think: <blockquote> “@ becomes € when you add Option.” </blockquote> Same key, just one extra modifier. <h4>Bonus: discover other hidden symbols</h4> macOS has many built-in characters behind the Option key: <ol> <li>Go to System Settings → Keyboard</li> <li>Enable Show Keyboard Viewer</li> <li>Hold Option or Option + Shift</li> </ol> You’ll see every hidden character live. Uniqlo cycling base layer https://danieljwilson.com/writing/tips-uniqlo-base-layer-cycling/ 2026-02-04T00:00:00.000Z As a cyclist sometimes I wear a base layer on cooler days. Base layers, as with much in cycling, can seem pretty overpriced. <img src="/assets/IMG-20260204155938224.png" alt="IMG-20260204155938224" width="250" loading="lazy"> The other day I wore one of my new long sleeve uniqlo t-shirts on a ride and it was great. Comfortable fit, soft and warm fabric. And 15€. <img src="/assets/IMG-20260204155938240.png" alt="IMG-20260204155938240" width="250" loading="lazy"> Strong recommend.