<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: Roberto Vázquez González</title>
    <description>The latest articles on DEV Community by Roberto Vázquez González (@robertovg).</description>
    <link>https://dev.to/robertovg</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F340951%2Fb84f70a1-e17f-496c-a770-df568da19aea.jpeg</url>
      <title>DEV Community: Roberto Vázquez González</title>
      <link>https://dev.to/robertovg</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/robertovg"/>
    <language>en</language>
    <item>
      <title>Introducing StrongList</title>
      <dc:creator>Roberto Vázquez González</dc:creator>
      <pubDate>Sat, 07 Nov 2026 09:29:36 +0000</pubDate>
      <link>https://dev.to/robertovg/introducing-stronglist-5g9o</link>
      <guid>https://dev.to/robertovg/introducing-stronglist-5g9o</guid>
      <description>&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fraw.githubusercontent.com%2Frobertovg%2FrobertovgGatsbyjs%2Fmain%2Fsrc%2Fposts%2Fimages%2Fstronglist.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fraw.githubusercontent.com%2Frobertovg%2FrobertovgGatsbyjs%2Fmain%2Fsrc%2Fposts%2Fimages%2Fstronglist.png" alt="StrongList header: Out now on iPhone &amp;amp; iPad" width="800" height="450"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;StrongList is out on the &lt;a href="https://apps.apple.com/app/stronglist/id6819074489" rel="noopener noreferrer"&gt;App Store&lt;/a&gt;. It's a workout log for iPhone and iPad, built with SwiftUI.&lt;/p&gt;

&lt;p&gt;I wanted to keep the scope straightforward: prepare a plan, record what you actually did and review it afterwards. The app has two tabs, Train and Progress, with a notebook-like interface.&lt;/p&gt;

&lt;h2&gt;
  
  
  What it does
&lt;/h2&gt;

&lt;p&gt;You can build plans from the exercise catalog or add your own exercises. For strength work, you log sets, reps and weight in kg or lb. Cardio and mobility use time, with optional distance.&lt;/p&gt;

&lt;p&gt;If you train at more than one gym, StrongList keeps machine numbers for each gym and exercise. You can also edit a plan during a workout, including adding or reordering exercises.&lt;/p&gt;

&lt;p&gt;If your trainer gives you a routine on paper, you can bring it in from a photo. StrongList gives you the instructions to use with the AI app you already have, like ChatGPT: you send it the photo, paste its reply back into StrongList, and the routine becomes a plan. StrongList itself doesn't send your photo or your data anywhere.&lt;/p&gt;

&lt;p&gt;While a workout is running, a Live Activity shows elapsed time, the completed exercise count and the next exercise on the Lock Screen. Supported iPhones also show workout information in the Dynamic Island.&lt;/p&gt;

&lt;p&gt;Progress lets you look at a session, week or month, including total sets, weight lifted and sets per muscle region. The app supports English and Spanish.&lt;/p&gt;

&lt;h2&gt;
  
  
  Price
&lt;/h2&gt;

&lt;p&gt;The launch price is &lt;strong&gt;$0.99 / €0.99&lt;/strong&gt; through November 20. From November 21, the regular price is &lt;strong&gt;$4.99 / €5.99&lt;/strong&gt;. Both are a one-time purchase.&lt;/p&gt;

&lt;p&gt;There are no subscriptions, ads or in-app purchases. You don't need a StrongList account either.&lt;/p&gt;

&lt;h2&gt;
  
  
  Where your data goes
&lt;/h2&gt;

&lt;p&gt;Your training records are stored locally on your device. iCloud Drive sync is on by default and can be turned off in Settings. If you use it, your training records are also saved as readable JSON files in your own iCloud Drive. There is no backend of mine storing your training history.&lt;/p&gt;

&lt;p&gt;That's the scope I've chosen for StrongList: following a plan and keeping a record of your workouts, rather than telling you how to train.&lt;/p&gt;

&lt;p&gt;If you try it, I'd like to hear what works and what doesn't. You can also follow StrongList on &lt;a href="https://x.com/stronglist_app" rel="noopener noreferrer"&gt;X&lt;/a&gt; or &lt;a href="https://www.instagram.com/stronglist_app/" rel="noopener noreferrer"&gt;Instagram&lt;/a&gt;.&lt;/p&gt;

</description>
      <category>showdev</category>
      <category>ios</category>
      <category>swiftui</category>
      <category>buildinpublic</category>
    </item>
    <item>
      <title>I Don't Need an Integration for Everything</title>
      <dc:creator>Roberto Vázquez González</dc:creator>
      <pubDate>Mon, 02 Nov 2026 11:46:10 +0000</pubDate>
      <link>https://dev.to/robertovg/i-dont-need-an-integration-for-everything-28b3</link>
      <guid>https://dev.to/robertovg/i-dont-need-an-integration-for-everything-28b3</guid>
      <description>&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F2fakc5i8wzpheajytb9a.jpg" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F2fakc5i8wzpheajytb9a.jpg" alt="Close-up shot of a person's hand flipping through pages of paper" width="800" height="450"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;I use MCPs, integrations and skills a lot in my AI workflow. They are incredibly useful. Giving an agent the ability to query a repository, inspect a project, search documentation or interact with another service can turn a simple chat into something much more powerful.&lt;/p&gt;

&lt;p&gt;But lately I've also been leaning on something much simpler: small browser snippets that extract the content I am already looking at and put it in my clipboard. No integration, no API, no OAuth flow, no server with access to my account.&lt;/p&gt;

&lt;p&gt;Just:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight plaintext"&gt;&lt;code&gt;I can see it → extract it → paste it into the AI
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;For some of the things I do, that's exactly the interface I want.&lt;/p&gt;

&lt;h2&gt;
  
  
  Two snippets
&lt;/h2&gt;

&lt;p&gt;In &lt;a href="https://robertovg.com/blog/verifiable-bookmarklets/" rel="noopener noreferrer"&gt;Verifiable Bookmarklets&lt;/a&gt; I listed the bookmarklets I use every day and said I would expand on one of them in a following post. This is that post. Two of them are built exactly for this workflow.&lt;/p&gt;

&lt;h3&gt;
  
  
  📤 Full Site To Markdown Extractor v1.0.3
&lt;/h3&gt;

&lt;p&gt;It takes the page I'm currently viewing and converts its relevant HTML content into Markdown. That might sound almost too simple, but Markdown is a fantastic format for feeding content to an AI.&lt;/p&gt;

&lt;p&gt;If I'm reading an article, a documentation page, an internal tool or any other page I want to discuss with an AI, I don't necessarily need an integration that knows how to access that website. I already have access to it. I can see the page. So I extract it, copy the Markdown and paste it into whatever model I want to use.&lt;/p&gt;

&lt;p&gt;If I need more than one page, it can also bundle other pages from the same site that I pick into a single Markdown document, fetched with the session I already have open. Those requests go through my browser, so they are still subject to same-origin rules and the rest of the browser's security restrictions. It doesn't bypass authentication or grant any new access: it can only fetch what I could already open myself.&lt;/p&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fy01ik03d6uba9z5tvuh7.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fy01ik03d6uba9z5tvuh7.png" alt="Full Site To Markdown Extractor running on robertovg.com, with the dialog to pick which pages to export" width="800" height="520"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;The source is &lt;a href="https://gist.github.com/robertovg/c6ed0ef05a3d1d4c958f551c2abb54c8" rel="noopener noreferrer"&gt;public as a Gist&lt;/a&gt;, and you can &lt;a href="https://robertovg.com/projects/verifiable-bookmarklets/?src=https%3A%2F%2Fgist.githubusercontent.com%2Frobertovg%2Fc6ed0ef05a3d1d4c958f551c2abb54c8%2Fraw%2F0de7d7616583305f1aad920910e922e86c0e266d%2FFull%252520Site%252520To%252520Markdown%252520Extractor%252520-%252520bookmarklet.js" rel="noopener noreferrer"&gt;generate the bookmarklet from it with Verifiable Bookmarklets&lt;/a&gt;.&lt;/p&gt;

&lt;h3&gt;
  
  
  🐙 GitHub PR Feedback Extractor v1.0.0
&lt;/h3&gt;

&lt;p&gt;The second one is more specific. When I'm looking at a GitHub pull request, it collects the review comments on the page and opens a dialog with them grouped by author. I tick the ones I care about and it copies them to my clipboard as a Markdown checklist, keeping the file path and line range of each comment, plus any suggested changes as &lt;code&gt;diff&lt;/code&gt; blocks. Comments from automated reviewers, like Copilot, can also be filtered by severity.&lt;/p&gt;

&lt;p&gt;It only extracts the comments currently rendered on the page and that I select. If a thread is collapsed or not loaded, it isn't included. It structures the feedback, but it doesn't evaluate whether a comment is correct or relevant. That part is still up to me, or to what I ask the model.&lt;/p&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F9kgee2s2jqk9emyyijqt.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F9kgee2s2jqk9emyyijqt.png" alt="GitHub PR Feedback Extractor on a StrongList pull request, with Copilot's review comments and my replies grouped by author" width="800" height="520"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;PR discussions are exactly the kind of context that makes an AI answer much more useful. Instead of explaining the whole conversation manually, I copy the relevant feedback and ask the model what I want:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;"Here's the PR feedback. What do you think?"&lt;/li&gt;
&lt;li&gt;"What would you change?"&lt;/li&gt;
&lt;li&gt;"Are these comments technically correct?"&lt;/li&gt;
&lt;li&gt;"Help me address these issues."&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;And again, the important part is that the AI doesn't need GitHub access. I am giving it the content I want it to see. Not needing GitHub API credentials doesn't make a private PR public or automatically safe to share, though. Whatever I select still gets sent to the AI provider I paste it into.&lt;/p&gt;

&lt;p&gt;The source is also &lt;a href="https://gist.github.com/robertovg/27b8a3da1dce1659239a0dfd26a671e5" rel="noopener noreferrer"&gt;public as a Gist&lt;/a&gt;, and you can &lt;a href="https://robertovg.com/projects/verifiable-bookmarklets/?src=https%3A%2F%2Fgist.githubusercontent.com%2Frobertovg%2F27b8a3da1dce1659239a0dfd26a671e5%2Fraw%2F5943bbc08fdccb79daa21063a12dd20495db81d5%2FGitHub%252520PR%252520Feedback%252520Extractor%252520-%252520bookmarklet.js" rel="noopener noreferrer"&gt;generate the bookmarklet from it with Verifiable Bookmarklets&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  Client versus server
&lt;/h2&gt;

&lt;p&gt;There is a subtle architectural difference here that I think is worth making explicit.&lt;/p&gt;

&lt;p&gt;A bookmarklet runs on the client. The page is already loaded in my browser, and the browser has access to the content because I have access to it. The snippet operates on that page, produces some output, and then I decide what happens next:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight plaintext"&gt;&lt;code&gt;Private page
     ↓
My browser
     ↓
Bookmarklet
     ↓
Markdown in clipboard
     ↓
AI prompt
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;The AI provider doesn't need credentials for the original service, an API token, or an integration capable of going back to the source. It simply receives the text I explicitly pasted into the conversation.&lt;/p&gt;

&lt;p&gt;An integration changes the architecture:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight plaintext"&gt;&lt;code&gt;Private service
     ↓
Integration / MCP server
     ↓
AI agent
     ↓
Requests more data when it wants it
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;That second model can obviously be extremely powerful. But it also creates a different security boundary. That doesn't necessarily mean unrestricted or autonomous access: the host decides which tools are exposed and with which permissions, and calls may need my approval or be restricted. Still, within those limits, the agent can go back to the source by itself.&lt;/p&gt;

&lt;h2&gt;
  
  
  Seeing something doesn't mean an AI should be able to retrieve it
&lt;/h2&gt;

&lt;p&gt;This is the main idea behind these snippets. If I can access some content, that doesn't mean I should grant an AI-controlled server access to my personal account so it can retrieve it. Those are two completely different permissions.&lt;/p&gt;

&lt;p&gt;Imagine I am looking at a private GitHub PR. I have access to the repository. That doesn't automatically mean I want an AI agent to have a GitHub token that lets it browse my repositories whenever it wants.&lt;/p&gt;

&lt;p&gt;Or my email. I can open my inbox and read an email. That doesn't mean I want an AI integration with permission to search my entire mailbox.&lt;/p&gt;

&lt;p&gt;The same applies to other personal services and accounts. An integration might give an agent access to things such as:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;private repository contents&lt;/li&gt;
&lt;li&gt;pull requests and issues&lt;/li&gt;
&lt;li&gt;emails and attachments&lt;/li&gt;
&lt;li&gt;contacts&lt;/li&gt;
&lt;li&gt;calendars&lt;/li&gt;
&lt;li&gt;documents&lt;/li&gt;
&lt;li&gt;personal account information&lt;/li&gt;
&lt;li&gt;financial information&lt;/li&gt;
&lt;li&gt;other private data exposed through an API&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;There is an important difference between these two statements:&lt;/p&gt;

&lt;blockquote&gt;
&lt;p&gt;"The AI can access my GitHub account."&lt;/p&gt;

&lt;p&gt;"I gave the AI the contents of this PR."&lt;/p&gt;
&lt;/blockquote&gt;

&lt;p&gt;The second is a much more explicit boundary. I decide what leaves the source system. I can inspect it before sending it, remove something sensitive, and choose the provider and the model. And when the task is finished, the AI doesn't have a credential that can later be used to retrieve more data.&lt;/p&gt;

&lt;p&gt;That's not perfect security. Once I paste something into an AI service, I've obviously shared that content with that service, and its data-handling policies still matter. But the scope of the permission is different. &lt;strong&gt;I am sharing data rather than granting access.&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;So the important question isn't simply:&lt;/p&gt;

&lt;blockquote&gt;
&lt;p&gt;"Can the AI help me if it has access?"&lt;/p&gt;
&lt;/blockquote&gt;

&lt;p&gt;Of course it can. The question is:&lt;/p&gt;

&lt;blockquote&gt;
&lt;p&gt;"Does this particular task justify giving the AI the ability to retrieve that data itself?"&lt;/p&gt;
&lt;/blockquote&gt;

&lt;p&gt;Sometimes the answer is yes. Sometimes it isn't. And I like having both options.&lt;/p&gt;

&lt;h2&gt;
  
  
  The strange thing about HTML to Markdown
&lt;/h2&gt;

&lt;p&gt;This is also why the Full Site To Markdown Extractor feels particularly interesting to me. If the browser already has an HTML document, converting it into Markdown is fundamentally a client-side transformation. So requiring a backend just to do this feels strange:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight plaintext"&gt;&lt;code&gt;HTML → Markdown
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;A backend would mean sending the page somewhere else. That doesn't mean every backend stores or logs it, but now you have another system involved, another trust boundary, and another component whose security and data-handling practices you have to trust.&lt;/p&gt;

&lt;p&gt;For a public webpage, perhaps that doesn't matter much. For a private one, it matters considerably more. And the browser already has everything required to perform the transformation, so why send it somewhere else?&lt;/p&gt;

&lt;p&gt;The simplest architecture is sometimes also the one with the smallest attack surface.&lt;/p&gt;

&lt;h2&gt;
  
  
  This isn't anti-integration
&lt;/h2&gt;

&lt;p&gt;I'm not arguing that integrations or MCPs are bad. Quite the opposite: I use them extensively.&lt;/p&gt;

&lt;p&gt;If I am working on a project and I want an agent to inspect the repository, understand the codebase, make changes, run tests and iterate, giving it the appropriate tooling can be incredibly productive. There are tasks where persistent, programmatic access is exactly what you want.&lt;/p&gt;

&lt;p&gt;But there are also tasks where I don't need an agent to have access. I just need to give it some context. That's a much smaller permission.&lt;/p&gt;

&lt;h2&gt;
  
  
  The clipboard becomes the bridge
&lt;/h2&gt;

&lt;p&gt;This is where my clipboard history becomes surprisingly useful. I basically treat it as a temporary context buffer. One caveat: clipboard managers can keep or synchronize what I copy, sensitive content included, so it's worth knowing where that history is stored.&lt;/p&gt;

&lt;p&gt;I extract something once, and then I can try it with ChatGPT, or Claude, or another model, or come back to it later:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight plaintext"&gt;&lt;code&gt;Private page
      ↓
   Extract
      ↓
   Clipboard
      ↓
 ┌────┼─────┐
 ↓    ↓     ↓
AI A  AI B  AI C
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Each model receives only the content I explicitly provided. None of them needs credentials for the original service. The source service doesn't need to know which AI I'm using, and I don't have to decide which provider deserves permanent access to my account. I simply choose what context crosses the boundary.&lt;/p&gt;

&lt;p&gt;It also makes experimenting with different models much easier. I don't have to install an integration for every provider, give every provider access to my accounts, or build a complicated data pipeline for something that ultimately amounts to:&lt;/p&gt;

&lt;blockquote&gt;
&lt;p&gt;"Here is some context. Please help me with it."&lt;/p&gt;
&lt;/blockquote&gt;

&lt;h2&gt;
  
  
  Sometimes the boring solution wins
&lt;/h2&gt;

&lt;p&gt;There is a tendency in AI tooling to solve everything by adding another integration. Need context from GitHub? Add GitHub. From Notion? Add Notion. From email? Add Gmail. From some internal application? Build an MCP server.&lt;/p&gt;

&lt;p&gt;And sometimes that's absolutely the right solution. But sometimes the entire requirement is:&lt;/p&gt;

&lt;blockquote&gt;
&lt;p&gt;"I'm looking at this thing and I want an AI to understand it."&lt;/p&gt;
&lt;/blockquote&gt;

&lt;p&gt;In that situation, a bookmarklet that converts the page to Markdown and puts it in my clipboard can be a surprisingly good piece of infrastructure:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;It's local.&lt;/li&gt;
&lt;li&gt;It's explicit.&lt;/li&gt;
&lt;li&gt;It's composable.&lt;/li&gt;
&lt;li&gt;It works with different AI providers.&lt;/li&gt;
&lt;li&gt;It doesn't require another backend.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;And, most importantly, it doesn't turn "I can access this" into "an AI agent can access this". That distinction is becoming increasingly important as AI agents get more capable.&lt;/p&gt;

&lt;p&gt;For me, these snippets aren't replacing MCPs or integrations. They're another permission model. When I want an agent to operate on my data, I give it the appropriate integration. When I just want to give an AI some context, I copy the context.&lt;/p&gt;

&lt;p&gt;Sometimes the best integration is still Ctrl+C, Ctrl+V.&lt;/p&gt;

&lt;p&gt;Image by Pexels user &lt;a href="https://www.pexels.com/photo/close-up-shot-of-a-person-s-hand-flipping-through-pages-of-paper-7054418/" rel="noopener noreferrer"&gt;Kindel Media&lt;/a&gt; used under the Pexels license.&lt;/p&gt;

</description>
      <category>ai</category>
      <category>mcp</category>
      <category>bookmarklets</category>
      <category>privacy</category>
    </item>
    <item>
      <title>Verifiable Bookmarklets</title>
      <dc:creator>Roberto Vázquez González</dc:creator>
      <pubDate>Mon, 27 Jul 2026 00:00:00 +0000</pubDate>
      <link>https://dev.to/robertovg/verifiable-bookmarklets-ae9</link>
      <guid>https://dev.to/robertovg/verifiable-bookmarklets-ae9</guid>
      <description>&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fx8d0hhhdl0wcztt1q5xr.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fx8d0hhhdl0wcztt1q5xr.png" alt="Browser dragging bookmarklet to the bookmark" width="800" height="500"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;I just added a small new tool to the site: &lt;a href="https://robertovg.com/projects/verifiable-bookmarklets/" rel="noopener noreferrer"&gt;Verifiable Bookmarklets&lt;/a&gt;. Every bookmarklet it generates carries a SHA-256 fingerprint, so anyone can check that the &lt;code&gt;javascript:&lt;/code&gt; link they received was really built from the source code they were shown.&lt;/p&gt;

&lt;p&gt;It is a lightweight reproducible-builds mindset applied to bookmarklets. Nothing more than that.&lt;/p&gt;

&lt;p&gt;The question it answers is:&lt;/p&gt;

&lt;blockquote&gt;
&lt;p&gt;Can this bookmarklet be proven to match this source code?&lt;/p&gt;
&lt;/blockquote&gt;

&lt;p&gt;That is all. It does not make bookmarklets secure. It does not sandbox them. It does not replace browser extensions or browser stores. A verified bookmarklet can still be bad code, dangerous code or code you should not run. But at least you can know whether the bookmarklet you received is exactly the bookmarklet that was built from the source being shown.&lt;/p&gt;

&lt;h2&gt;
  
  
  Why bookmarklets?
&lt;/h2&gt;

&lt;p&gt;Bookmarklets are old, simple and still useful. You drag a link to the bookmarks bar, click it later and the browser runs a small piece of JavaScript in the current page. No install flow, no store review, no account, no backend.&lt;/p&gt;

&lt;p&gt;I still use them every day at work. These are real ones I have written over the years to make repetitive tasks easier, all of them from a pre-AI world, but the same principles apply:&lt;/p&gt;

&lt;h3&gt;
  
  
  🐙 GitHub PR Feedback Extractor v1.0.0
&lt;/h3&gt;

&lt;p&gt;The newest one. Run it on a PR and it opens a dialog listing every comment grouped by author. I tick the ones I care about and it copies a Markdown extraction I can paste straight into an AI agent to iterate on and address.&lt;/p&gt;

&lt;p&gt;A GitHub skill with access to the GitHub CLI could do the same, but when I already have the comments in front of me, clicking a bookmarklet is just faster.&lt;/p&gt;

&lt;h3&gt;
  
  
  🔗📋 Copy page link v1.0.3
&lt;/h3&gt;

&lt;p&gt;Generates a Markdown link to the current page and copies it to the clipboard. If there is text selected it becomes the link text, otherwise it falls back to the page title.&lt;/p&gt;

&lt;h3&gt;
  
  
  📤 Full Site To Markdown Extractor v1.0.3
&lt;/h3&gt;

&lt;p&gt;Extracts the content of a page as Markdown so I can feed it to an AI agent. I will expand on this one in a following post.&lt;/p&gt;

&lt;h3&gt;
  
  
  🕳 🧩 🦑 📋 Show / Hide the Content Extractor v1.0.9
&lt;/h3&gt;

&lt;p&gt;Shows an input on the page where I can type a CSS selector. It highlights the matching elements and copies their content to the clipboard.&lt;/p&gt;

&lt;p&gt;Most of my others are too specific to my daily work to be worth publishing. The point is the shape of the list: tiny, boring, single-purpose automations that pull an identifier out of a page, reformat something into Markdown, or save me four clicks in an internal tool. An extension would be overkill for every single one of them.&lt;/p&gt;

&lt;p&gt;Notice the version numbers too. These things get edited, regenerated and re-dragged constantly, which is exactly where provenance starts to matter.&lt;/p&gt;

&lt;h2&gt;
  
  
  Sharing them never worked well
&lt;/h2&gt;

&lt;p&gt;There is no standard way to share a bookmarklet. Drag one to your desktop and you get a &lt;code&gt;.inetloc&lt;/code&gt; file on macOS, a &lt;code&gt;.url&lt;/code&gt; file on Windows or a &lt;code&gt;.desktop&lt;/code&gt; file on Linux. They all contain the &lt;code&gt;javascript:&lt;/code&gt; URL, but they are not human-readable and they are not nice to share.&lt;/p&gt;

&lt;p&gt;For years I generated bookmarklets with &lt;a href="https://caiorss.github.io/bookmarklet-maker/" rel="noopener noreferrer"&gt;Bookmarklet Maker&lt;/a&gt;, which takes normal JavaScript and gives you back a bookmarklet ready to drag into the browser. That solved generation well enough that I never looked for anything else. I shared them with colleagues, they shared them with me, and I treated them as informal web scripts without any more pretension than that.&lt;/p&gt;

&lt;h2&gt;
  
  
  The part that changed recently
&lt;/h2&gt;

&lt;p&gt;Because of LLMs, the generation step has become trivial. Ask for "a bookmarklet that does X" and you get working code back in seconds. That is genuinely useful and I do it often. It also means a lot more &lt;code&gt;javascript:&lt;/code&gt; URLs are being generated, pasted into Slack and shared between people who never saw the source.&lt;/p&gt;

&lt;p&gt;So more mates share bookmarklets with me, and I share more with them. Generation is no longer the bottleneck.&lt;/p&gt;

&lt;p&gt;Which is when it hit me: &lt;strong&gt;how do I know that the bookmarklet I received is exactly the one that was built from the source I was shown?&lt;/strong&gt; And how can I be sure the bookmarklet I am about to share is exactly the one built from the source I am showing?&lt;/p&gt;

&lt;p&gt;Whether that source was written by me, by a colleague or by a model does not really change the question:&lt;/p&gt;

&lt;blockquote&gt;
&lt;p&gt;Is this bookmarklet exactly what this source produces?&lt;/p&gt;
&lt;/blockquote&gt;

&lt;h3&gt;
  
  
  Before
&lt;/h3&gt;

&lt;ul&gt;
&lt;li&gt;Share an opaque &lt;code&gt;javascript:&lt;/code&gt; URL&lt;/li&gt;
&lt;li&gt;Trust the author&lt;/li&gt;
&lt;li&gt;Impossible to compare&lt;/li&gt;
&lt;li&gt;"Looks fine"&lt;/li&gt;
&lt;/ul&gt;

&lt;h3&gt;
  
  
  After
&lt;/h3&gt;

&lt;ul&gt;
&lt;li&gt;Share a source URL&lt;/li&gt;
&lt;li&gt;Verify the build&lt;/li&gt;
&lt;li&gt;Deterministic SHA-256&lt;/li&gt;
&lt;li&gt;"Matches source"&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  How it works
&lt;/h2&gt;

&lt;p&gt;Each generated bookmarklet contains a small metadata comment:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight javascript"&gt;&lt;code&gt;&lt;span class="cm"&gt;/*bm:v1;sha=&amp;lt;sha256&amp;gt;;src=&amp;lt;optional-url&amp;gt;*/&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;The tool builds the payload, hashes it with SHA-256 and embeds that comment. The hash does not include the metadata itself, which avoids the circular problem of hashing something that contains its own hash.&lt;/p&gt;

&lt;p&gt;The pipeline is deliberately short:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight plaintext"&gt;&lt;code&gt;normalize source (normalize newlines, trim)
      ↓
build payload
      ↓
SHA-256(payload)
      ↓
embed metadata
      ↓
share
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;The tool can already:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;generate a bookmarklet from pasted JavaScript source&lt;/li&gt;
&lt;li&gt;generate a bookmarklet from a raw JavaScript URL&lt;/li&gt;
&lt;li&gt;create a draggable bookmarklet link&lt;/li&gt;
&lt;li&gt;inspect a pasted or dropped bookmarklet&lt;/li&gt;
&lt;li&gt;show decoded JavaScript and metadata&lt;/li&gt;
&lt;li&gt;verify a bookmarklet against the current source&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Everything runs in the browser. No backend, no login, no publishing service.&lt;/p&gt;

&lt;p&gt;It also generates a shareable link to the tool with the source URL as a parameter, so anyone can review the source and generate the bookmarklet themselves:&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;&lt;a href="https://robertovg.com/projects/verifiable-bookmarklets/?src=&lt;url-encoded-source" rel="noopener noreferrer"&gt;https://robertovg.com/projects/verifiable-bookmarklets/?src=&amp;amp;lt;url-encoded-source&lt;/a&gt;&amp;gt;&lt;/strong&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  Two workflows
&lt;/h2&gt;

&lt;p&gt;That source URL parameter turned out to be the better way to share a bookmarklet.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Authoring.&lt;/strong&gt; Publish the source, for example as a GitHub Gist, and link to Verifiable Bookmarklets with that URL as a parameter. The tool fetches the source, generates the bookmarklet locally and shows the fingerprint. Every reader reviews plain JavaScript and builds the executable in their own browser instead of trusting a long opaque link. Pinning a specific Gist revision rather than the mutable latest URL makes the reference stronger, because the source itself cannot change under you.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Verification.&lt;/strong&gt; When someone shares the bookmarklet itself — "I found this in Slack, is it still the original?" — the embedded metadata is what helps. Inspect it, read the &lt;code&gt;src&lt;/code&gt; and &lt;code&gt;sha&lt;/code&gt;, fetch the source, regenerate and compare.&lt;/p&gt;

&lt;h2&gt;
  
  
  What verification means
&lt;/h2&gt;

&lt;p&gt;Verification means:&lt;/p&gt;

&lt;blockquote&gt;
&lt;p&gt;This bookmarklet matches this source.&lt;/p&gt;
&lt;/blockquote&gt;

&lt;p&gt;It does not mean:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;this bookmarklet is safe&lt;/li&gt;
&lt;li&gt;this bookmarklet is trustworthy&lt;/li&gt;
&lt;li&gt;this source is the original source&lt;/li&gt;
&lt;li&gt;this source URL will always return the same code&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;This is about provenance, not security. I think that distinction matters. Small tools become more useful when they make narrow claims and keep those claims honest.&lt;/p&gt;

&lt;h2&gt;
  
  
  The three functions that carry the idea
&lt;/h2&gt;

&lt;p&gt;Generating the payload and the fingerprint &lt;code&gt;createPayload&lt;/code&gt;, &lt;code&gt;createMetadata&lt;/code&gt; and &lt;code&gt;verify&lt;/code&gt;&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight javascript"&gt;&lt;code&gt;&lt;span class="kd"&gt;function&lt;/span&gt; &lt;span class="nf"&gt;normalizeSource&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nx"&gt;source&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;
  &lt;span class="k"&gt;return&lt;/span&gt; &lt;span class="nx"&gt;source&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;replace&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="sr"&gt;/&lt;/span&gt;&lt;span class="se"&gt;\r\n?&lt;/span&gt;&lt;span class="sr"&gt;/g&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="dl"&gt;'&lt;/span&gt;&lt;span class="se"&gt;\n&lt;/span&gt;&lt;span class="dl"&gt;'&lt;/span&gt;&lt;span class="p"&gt;).&lt;/span&gt;&lt;span class="nf"&gt;trim&lt;/span&gt;&lt;span class="p"&gt;()&lt;/span&gt;
&lt;span class="p"&gt;}&lt;/span&gt;

&lt;span class="kd"&gt;function&lt;/span&gt; &lt;span class="nf"&gt;minifySource&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nx"&gt;source&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;
  &lt;span class="k"&gt;return&lt;/span&gt; &lt;span class="nf"&gt;normalizeSource&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nx"&gt;source&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;
&lt;span class="p"&gt;}&lt;/span&gt;

&lt;span class="kd"&gt;function&lt;/span&gt; &lt;span class="nf"&gt;encodeBookmarkletValue&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nx"&gt;value&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;
  &lt;span class="k"&gt;return&lt;/span&gt; &lt;span class="nf"&gt;encodeURIComponent&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nx"&gt;value&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;
    &lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;replace&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="sr"&gt;/'/g&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="dl"&gt;'&lt;/span&gt;&lt;span class="s1"&gt;%27&lt;/span&gt;&lt;span class="dl"&gt;'&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;
    &lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;replace&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="sr"&gt;/"/g&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="dl"&gt;'&lt;/span&gt;&lt;span class="s1"&gt;%22&lt;/span&gt;&lt;span class="dl"&gt;'&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;
&lt;span class="p"&gt;}&lt;/span&gt;

&lt;span class="kd"&gt;function&lt;/span&gt; &lt;span class="nf"&gt;encodeBookmarkletBody&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nx"&gt;body&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;
  &lt;span class="k"&gt;return&lt;/span&gt; &lt;span class="s2"&gt;`javascript:&lt;/span&gt;&lt;span class="p"&gt;${&lt;/span&gt;&lt;span class="nf"&gt;encodeURIComponent&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nx"&gt;body&lt;/span&gt;&lt;span class="p"&gt;)}&lt;/span&gt;&lt;span class="s2"&gt;`&lt;/span&gt;
&lt;span class="p"&gt;}&lt;/span&gt;

&lt;span class="kd"&gt;function&lt;/span&gt; &lt;span class="nf"&gt;createPayload&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nx"&gt;source&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;
  &lt;span class="k"&gt;return&lt;/span&gt; &lt;span class="nf"&gt;encodeBookmarkletBody&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="s2"&gt;`(()=&amp;gt;{&lt;/span&gt;&lt;span class="p"&gt;${&lt;/span&gt;&lt;span class="nf"&gt;minifySource&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nx"&gt;source&lt;/span&gt;&lt;span class="p"&gt;)}&lt;/span&gt;&lt;span class="s2"&gt;})()`&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;
&lt;span class="p"&gt;}&lt;/span&gt;

&lt;span class="kd"&gt;function&lt;/span&gt; &lt;span class="nf"&gt;createMetadata&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nx"&gt;sha&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nx"&gt;src&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;
  &lt;span class="k"&gt;return&lt;/span&gt; &lt;span class="nx"&gt;src&lt;/span&gt; &lt;span class="p"&gt;?&lt;/span&gt; &lt;span class="s2"&gt;`/*bm:v1;sha=&lt;/span&gt;&lt;span class="p"&gt;${&lt;/span&gt;&lt;span class="nx"&gt;sha&lt;/span&gt;&lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="s2"&gt;;src=&lt;/span&gt;&lt;span class="p"&gt;${&lt;/span&gt;&lt;span class="nf"&gt;encodeBookmarkletValue&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nx"&gt;src&lt;/span&gt;&lt;span class="p"&gt;)}&lt;/span&gt;&lt;span class="s2"&gt;*/`&lt;/span&gt; &lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;`/*bm:v1;sha=&lt;/span&gt;&lt;span class="p"&gt;${&lt;/span&gt;&lt;span class="nx"&gt;sha&lt;/span&gt;&lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="s2"&gt;*/`&lt;/span&gt;
&lt;span class="p"&gt;}&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;And verifying, which rebuilds the payload from the decoded bookmarklet, hashes it and compares it against the embedded &lt;code&gt;sha&lt;/code&gt;. It never executes the code:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight javascript"&gt;&lt;code&gt;&lt;span class="k"&gt;async&lt;/span&gt; &lt;span class="kd"&gt;function&lt;/span&gt; &lt;span class="nf"&gt;verify&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nx"&gt;bookmarklet&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nx"&gt;source&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;
  &lt;span class="kd"&gt;const&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nx"&gt;decoded&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nx"&gt;hasScheme&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="nf"&gt;decodeBookmarklet&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nx"&gt;bookmarklet&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;
  &lt;span class="kd"&gt;const&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nx"&gt;metadata&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="nf"&gt;parseMetadata&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nx"&gt;decoded&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;
  &lt;span class="kd"&gt;const&lt;/span&gt; &lt;span class="nx"&gt;payloadBody&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="nf"&gt;removeMetadata&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nx"&gt;decoded&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;
  &lt;span class="kd"&gt;const&lt;/span&gt; &lt;span class="nx"&gt;payload&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="nx"&gt;hasScheme&lt;/span&gt; &lt;span class="p"&gt;?&lt;/span&gt; &lt;span class="nf"&gt;encodeBookmarkletBody&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nx"&gt;payloadBody&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="nx"&gt;payloadBody&lt;/span&gt;

  &lt;span class="k"&gt;return&lt;/span&gt; &lt;span class="nx"&gt;metadata&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nx"&gt;sha&lt;/span&gt; &lt;span class="o"&gt;===&lt;/span&gt; &lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="k"&gt;await&lt;/span&gt; &lt;span class="nf"&gt;sha256&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nf"&gt;createPayload&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nx"&gt;source&lt;/span&gt;&lt;span class="p"&gt;)))&lt;/span&gt;
&lt;span class="p"&gt;}&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;The deterministic step stays conservative on purpose: normalize newlines, trim the source, hash the generated payload. It just turns JavaScript you could run in your console into a bookmarklet you can check.&lt;/p&gt;

&lt;p&gt;The rest of the implementation lives in &lt;a href="https://github.com/robertovg/robertovgGatsbyjs/blob/main/src/pages/projects/verifiable-bookmarklets.js#L417" rel="noopener noreferrer"&gt;this site's GitHub repo&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  Try it
&lt;/h2&gt;

&lt;p&gt;&lt;a href="https://robertovg.com/projects/verifiable-bookmarklets/" rel="noopener noreferrer"&gt;Open Verifiable Bookmarklets&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;Instead of sending someone an opaque &lt;code&gt;javascript:&lt;/code&gt; URL, I can now send them a link to the source and let them generate the executable themselves, like this &lt;a href="https://robertovg.com/projects/verifiable-bookmarklets/?src=https%3A%2F%2Fgist.githubusercontent.com%2Frobertovg%2Fc6ed0ef05a3d1d4c958f551c2abb54c8%2Fraw%2F0de7d7616583305f1aad920910e922e86c0e266d%2FFull%252520Site%252520To%252520Markdown%252520Extractor%252520-%252520bookmarklet.js" rel="noopener noreferrer"&gt;📤 Full Site To Markdown Extractor&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;AI made generating bookmarklets trivial. I wanted to simplify the sharing process and give some transparency to the provenance of the code. This tool does that, and I hope it is useful to others too.&lt;/p&gt;

</description>
      <category>bookmarklets</category>
      <category>javascript</category>
      <category>tools</category>
      <category>provenance</category>
    </item>
    <item>
      <title>My /uses page, six years later</title>
      <dc:creator>Roberto Vázquez González</dc:creator>
      <pubDate>Wed, 15 Jul 2026 00:00:00 +0000</pubDate>
      <link>https://dev.to/robertovg/my-uses-page-six-years-later-1dn3</link>
      <guid>https://dev.to/robertovg/my-uses-page-six-years-later-1dn3</guid>
      <description>&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F6pz14gttsiz2q7vkmgv2.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F6pz14gttsiz2q7vkmgv2.png" alt="uses.tech banner 2026" width="800" height="439"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;I just updated my &lt;a href="https://robertovg.com/uses/" rel="noopener noreferrer"&gt;/uses page&lt;/a&gt; for the first time since... well, since a version of me that wrote JavaScript without types, deployed with create-react-app, and had never talked to an AI in a terminal.&lt;/p&gt;

&lt;p&gt;Instead of quietly rewriting it and pretending nothing happened, I turned the page into a small timeline: what I use now on top, and the old version preserved below as historical data. I like these pages more when they're honest about time passing.&lt;/p&gt;

&lt;h2&gt;
  
  
  What actually changed
&lt;/h2&gt;

&lt;p&gt;Looking at the diff between 2019-me and 2026-me, a few things stand out:&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;TypeScript won.&lt;/strong&gt; I don't write plain JavaScript anymore, anywhere. TS on the frontend, TS on the backend with Nest.js. Python and Go joined the toolbox too, and Java quietly became an old friend I never call.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;The tooling turned over almost completely.&lt;/strong&gt; Alfred became Raycast. My terminal became Warp. create-react-app became Vite. Gatsby and Next.js are still around in projects I maintain, but Astro is where my new sites go. And most of my new projects now live on Cloudflare (Pages, Workers, KV) instead of scattered across half the internet.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;AI agents are the biggest change by far.&lt;/strong&gt; In 2019, "Editors" was a section about VSCode plugins. Now it's "Editors / AI Agents", because Claude Code, Codex, Copilot and friends are part of my daily workflow, both at work and on side projects. I still care about the same thing I cared about six years ago, though: the best tools are the ones that make life easier without the effort behind them being visible. That hasn't changed, only the tools have.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;The hardware section got simpler.&lt;/strong&gt; Same setup at home and at &lt;a href="https://corkerspace.com" rel="noopener noreferrer"&gt;Corker&lt;/a&gt; (the coworking space I co-founded): MacBook Pro 16" behind an external monitor, external keyboard and mouse, so moving between the two feels like nothing. The AKG K240 + Aune DAC combo from the old list is still going strong. Some things are worth keeping.&lt;/p&gt;

&lt;h2&gt;
  
  
  What all of this is for
&lt;/h2&gt;

&lt;p&gt;A /uses page can look like a list of shiny things, but for me it's the workshop, not the work. Most of this stack is currently pointed at being productive. AI is here to stay, and I genuinely like how it simplifies code creation, both at work and outside of it. But I also like to keep creating things from scratch, no AI assistance, just to keep my coding brain sharp and in good shape, up to date with how things actually work under the hood.&lt;/p&gt;

&lt;p&gt;If something on the list catches your eye and you want to know more, reach out. And if you have a /uses page of your own, I'd genuinely love to see the diff between your past and present self.&lt;/p&gt;

</description>
      <category>website</category>
      <category>uses</category>
      <category>ai</category>
      <category>tools</category>
    </item>
    <item>
      <title>Rebooting this website</title>
      <dc:creator>Roberto Vázquez González</dc:creator>
      <pubDate>Sun, 05 Jul 2026 00:00:00 +0000</pubDate>
      <link>https://dev.to/robertovg/rebooting-this-website-388b</link>
      <guid>https://dev.to/robertovg/rebooting-this-website-388b</guid>
      <description>&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F15amfvh7y8ehbmqbmqwb.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F15amfvh7y8ehbmqbmqwb.png" alt="Shell session rebooting my site" width="800" height="533"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;It's been a while.&lt;/p&gt;

&lt;p&gt;&lt;a href="https://robertovg.com/blog/here-my-uses-page/" rel="noopener noreferrer"&gt;The last article&lt;/a&gt; I published here was back in 2020, when I wrote about my &lt;a href="https://robertovg.com/uses/" rel="noopener noreferrer"&gt;/uses&lt;/a&gt; page. Looking back, it's hard to believe how much has changed since then, both in my life and in the tools I use every day. Updating that page is definitely overdue, but I'll save that for another day.&lt;/p&gt;

&lt;p&gt;The world went through the COVID-19 pandemic and lockdowns, life moved on, and over the years this website slowly stopped being a priority.&lt;/p&gt;

&lt;p&gt;Not because I stopped building things, quite the opposite.&lt;/p&gt;

&lt;p&gt;Since then, I've been working at EverQuote, where I've had the opportunity to work on products at a much larger scale than ever before. I also married my life partner, Nuria, and not long afterwards our family started to grow. First came Berta, and two years later Elio, by far the biggest and best change in our lives.&lt;/p&gt;

&lt;p&gt;Like many people, I simply had different priorities.&lt;/p&gt;

&lt;h2&gt;
  
  
  Simplifying things
&lt;/h2&gt;

&lt;p&gt;One thing I did want to improve was where all my personal projects live.&lt;/p&gt;

&lt;p&gt;Over the years I ended up hosting different sites across AWS, Vercel, Netlify, Google Cloud and a few other places. It was fun, and a great excuse to learn and compare different platforms, but it also meant jumping between providers every time I wanted to touch one of my projects.&lt;/p&gt;

&lt;p&gt;Lately I've been moving everything to Cloudflare.&lt;/p&gt;

&lt;p&gt;For my personal projects, it hits a really nice balance between simplicity, performance and cost. Having hosting, deployments and domain management in the same place makes maintaining everything much easier.&lt;/p&gt;

&lt;p&gt;Migrating everything there also gave me the perfect excuse to revisit this website, refresh the content and make it better reflect who I am today.&lt;/p&gt;

&lt;h2&gt;
  
  
  Looking forward
&lt;/h2&gt;

&lt;p&gt;This isn't a "big relaunch". Just a small reboot.&lt;/p&gt;

&lt;p&gt;I have a personal project I'm very excited about and that I hope to start shipping soon.&lt;/p&gt;

&lt;p&gt;Its name is &lt;strong&gt;SongForm&lt;/strong&gt;.&lt;/p&gt;

&lt;p&gt;My hope is that this website becomes a place where I occasionally document that journey, share things I learn along the way, and publish the odd article when I have something worth sharing.&lt;/p&gt;

&lt;h2&gt;
  
  
  Why keep writing?
&lt;/h2&gt;

&lt;p&gt;I did wonder whether maintaining a personal website still made sense in a world increasingly filled with AI-generated content, or what better said "AI slop."&lt;/p&gt;

&lt;p&gt;The more I thought about it, the more I realised that's exactly why it does.&lt;/p&gt;

&lt;p&gt;I use AI every day, both personally and professionally. I think it's one of the most useful &lt;strong&gt;TOOLS&lt;/strong&gt; I've ever had. It helps me build faster, execute ideas more efficiently, research unfamiliar topics and augment my knowledge.&lt;/p&gt;

&lt;p&gt;That said, I've found it's most valuable when I already understand the subject well enough to review its output critically, separate facts from hallucinations and apply my own experience.&lt;/p&gt;

&lt;p&gt;Personal websites were never about producing as much content as possible. They're about documenting a journey, sharing lessons learned and occasionally building something that someone else might find useful.&lt;/p&gt;

&lt;p&gt;AI can help me write. It can help me code. But it can't accumulate my experiences for me.&lt;/p&gt;

&lt;p&gt;So here's to rebooting this little corner of the internet.&lt;/p&gt;

&lt;p&gt;Hopefully this is the start of writing here a little more often.&lt;/p&gt;

</description>
      <category>website</category>
      <category>cloudflare</category>
      <category>family</category>
      <category>ai</category>
    </item>
  </channel>
</rss>
