<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>bokstudio IA — News on AI applied to web development</title><description>A daily selection of news about artificial intelligence applied to web development: tools, models, agents, Astro, WordPress and frontend. Original summaries in English and Spanish.</description><link>https://ia.bokstudio.com/</link><language>en</language><item><title>What Claude Code, Codex and Mistral Vibe store in their session logs</title><link>https://ia.bokstudio.com/en/news/9efc6f9b-what-claude-code-codex-and-mistral-vibe-store-in-their-session-logs/</link><guid isPermaLink="true">https://ia.bokstudio.com/en/news/9efc6f9b-what-claude-code-codex-and-mistral-vibe-store-in-their-session-logs/</guid><description>All three coding agents write everything they do to disk, in undocumented formats. The author of Agent Standup, an open-source tool that reads those files to show which agent is working or waiting, explains where they live, which records matter and how sub-agents show up. Useful for anyone monitoring several sessions at once.</description><pubDate>Fri, 16 Oct 2026 12:50:35 GMT</pubDate><source url="https://dev.to/atoll92/-reading-the-undocumented-session-logs-of-claude-code-codex-and-mistral-vibe-4i67">DEV Community</source><category>Agents &amp; MCP</category><category>Tools</category></item><item><title>An experiment with 40 hidden bugs measures how much an AI finds as a QA tester</title><link>https://ia.bokstudio.com/en/news/ee8de698-an-experiment-with-40-hidden-bugs-measures-how-much-an-ai-finds-as-a-q/</link><guid isPermaLink="true">https://ia.bokstudio.com/en/news/ee8de698-an-experiment-with-40-hidden-bugs-measures-how-much-an-ai-finds-as-a-q/</guid><description>A developer hid 40 bugs in a test online store and had Claude Code explore it with Playwright using three prompt styles. With a full checklist it found about 30 bugs on average for under a dollar per run; five bugs were missed by every run. The takeaway: a useful first pass before release, not a replacement for human testers.</description><pubDate>Fri, 16 Oct 2026 11:06:18 GMT</pubDate><source url="https://dev.to/rodrinobrega/i-hid-40-bugs-in-an-online-store-and-let-an-ai-play-qa-tester-1if2">DEV Community</source><category>AI for coding</category><category>Frontend</category><category>Tools</category></item><item><title>Guide to self-hosting EmDash, Cloudflare’s Astro-based CMS, on a VPS</title><link>https://ia.bokstudio.com/en/news/c2a25e17-guide-to-self-hosting-emdash-cloudflare-s-astro-based-cms-on-a-vps/</link><guid isPermaLink="true">https://ia.bokstudio.com/en/news/c2a25e17-guide-to-self-hosting-emdash-cloudflare-s-astro-based-cms-on-a-vps/</guid><description>EmDash is an open-source CMS from Cloudflare built on Astro and designed for both humans and agents. This tutorial shows how to install it outside Cloudflare, on a Hetzner VPS with Node.js, SQLite, Caddy and PM2, making it an option worth weighing against WordPress for Astro projects.</description><pubDate>Thu, 15 Oct 2026 18:22:22 GMT</pubDate><source url="https://dev.to/mandrasch/quick-test-of-emdash-cms-deployment-on-hetzner-vps-nodejs-518e">DEV Community</source><category>Astro</category><category>Hosting &amp; deploy</category></item><item><title>Meta releases Muse Spark Open, an open-weight model built for coding</title><link>https://ia.bokstudio.com/en/news/499fcbd0-meta-releases-muse-spark-open-an-open-weight-model-built-for-coding/</link><guid isPermaLink="true">https://ia.bokstudio.com/en/news/499fcbd0-meta-releases-muse-spark-open-an-open-weight-model-built-for-coding/</guid><description>Meta has released the weights of Muse Spark Open, a multimodal model with about 816B parameters (42B active) and a one-million-token context, aimed at coding and long agentic tasks. It is on Hugging Face and available through providers such as Together AI, Fireworks and OpenRouter, although its license restricts commercial use.</description><pubDate>Thu, 15 Oct 2026 17:47:39 GMT</pubDate><source url="https://research.meta.ai/blog/muse-spark-open">Meta AI</source><category>Models</category><category>AI for coding</category><category>Agents &amp; MCP</category></item><item><title>Vercel Ship 2026: AI SDK 7, MCP events and more backend for agents</title><link>https://ia.bokstudio.com/en/news/d0ecc421-vercel-ship-2026-ai-sdk-7-mcp-events-and-more-backend-for-agents/</link><guid isPermaLink="true">https://ia.bokstudio.com/en/news/d0ecc421-vercel-ship-2026-ai-sdk-7-mcp-events-and-more-backend-for-agents/</guid><description>Vercel recaps its annual event, focused on infrastructure for agents: AI SDK 7 for building multi-turn, tool-using agents, MCP Events in beta to react to MCP server events, and new budget and privacy controls in AI Gateway. On the traditional side, Python becomes generally available and Daemons arrive for background workloads.</description><pubDate>Thu, 15 Oct 2026 16:00:00 GMT</pubDate><source url="https://vercel.com/blog/vercel-ship-2026-recap">Vercel</source><category>Agents &amp; MCP</category><category>Hosting &amp; deploy</category><category>Tools</category></item><item><title>Cloudflare introduces Void to deploy full-stack Vite apps on Workers</title><link>https://ia.bokstudio.com/en/news/c15eca6d-cloudflare-introduces-void-to-deploy-full-stack-vite-apps-on-workers/</link><guid isPermaLink="true">https://ia.bokstudio.com/en/news/c15eca6d-cloudflare-introduces-void-to-deploy-full-stack-vite-apps-on-workers/</guid><description>Void is an open-source SDK, CLI and platform that turns a Vite app into a full-stack Cloudflare Workers app, inferring and provisioning resources such as databases, queues and AI bindings. It integrates with Astro, React Router, SvelteKit and Nuxt, among others, and its CLI is designed so agents can debug, read logs and deploy.</description><pubDate>Thu, 15 Oct 2026 14:00:00 GMT</pubDate><source url="https://blog.cloudflare.com/deploy-with-void/">Cloudflare Blog</source><category>Hosting &amp; deploy</category><category>JavaScript</category><category>Agents &amp; MCP</category></item><item><title>Cloudflare and OpenAI test signals to help ChatGPT discover web content</title><link>https://ia.bokstudio.com/en/news/10f08c7e-cloudflare-and-openai-test-signals-to-help-chatgpt-discover-web-conten/</link><guid isPermaLink="true">https://ia.bokstudio.com/en/news/10f08c7e-cloudflare-and-openai-test-signals-to-help-chatgpt-discover-web-conten/</guid><description>A pilot between Cloudflare and OpenAI uses signals from Cloudflare’s network so ChatGPT search can find new pages, know when they change and skip wasted requests. It matters to site owners because it shapes how AI search discovers their content; Cloudflare says it relies on metadata, not page content for training, and access is still managed with AI Crawl Control.</description><pubDate>Thu, 15 Oct 2026 13:00:00 GMT</pubDate><source url="https://blog.cloudflare.com/cloudflare-openai-discovery-agentic-web/">Cloudflare Blog</source><category>AI &amp; SEO</category><category>Hosting &amp; deploy</category></item><item><title>Astro content collections, Zod and an agent to maintain a tools directory</title><link>https://ia.bokstudio.com/en/news/f6825120-astro-content-collections-zod-and-an-agent-to-maintain-a-tools-directo/</link><guid isPermaLink="true">https://ia.bokstudio.com/en/news/f6825120-astro-content-collections-zod-and-an-agent-to-maintain-a-tools-directo/</guid><description>The author of a directory of AI tools for developers explains how rankings are modeled as Zod-validated data in Astro content collections. The post also covers making the site readable by chatbots and letting an AI agent ship changes without bypassing the rules.</description><pubDate>Thu, 15 Oct 2026 08:38:11 GMT</pubDate><source url="https://dev.to/xavier_enciso_6d91f56d298/building-a-systematic-ai-tool-guide-for-developers-with-astro-content-collections-zod-and-an-ai-1m20">DEV Community</source><category>Astro</category><category>AI for coding</category></item><item><title>Open models now carry 71.5% of tokens on Vercel’s AI Gateway</title><link>https://ia.bokstudio.com/en/news/1bf1c0d6-open-models-now-carry-71-5-of-tokens-on-vercel-s-ai-gateway/</link><guid isPermaLink="true">https://ia.bokstudio.com/en/news/1bf1c0d6-open-models-now-carry-71-5-of-tokens-on-vercel-s-ai-gateway/</guid><description>Vercel’s monthly report on AI Gateway traffic shows that in September open-weight models handled 71.5% of tokens but only 23% of spend, with DeepSeek leading on volume. It is useful context when picking production models: price per token is falling, but total spend keeps growing.</description><pubDate>Thu, 15 Oct 2026 07:00:00 GMT</pubDate><source url="https://vercel.com/blog/ai-gateway-production-index-october-2026">Vercel</source><category>Models</category></item><item><title>Vercel Drives gives agents persistent storage in Vercel Sandbox</title><link>https://ia.bokstudio.com/en/news/dd6e11f6-vercel-drives-gives-agents-persistent-storage-in-vercel-sandbox/</link><guid isPermaLink="true">https://ia.bokstudio.com/en/news/dd6e11f6-vercel-drives-gives-agents-persistent-storage-in-vercel-sandbox/</guid><description>Vercel Drives decouples an agent’s files from the environment it runs in, so it can switch sandboxes and keep the same workspace. It is now available for Vercel Sandbox on all plans, and drives can be mounted by several sandboxes or forked to try alternatives in parallel.</description><pubDate>Thu, 15 Oct 2026 07:00:00 GMT</pubDate><source url="https://vercel.com/blog/introducing-vercel-drives-persistent-storage-for-agents">Vercel</source><category>Agents &amp; MCP</category><category>Hosting &amp; deploy</category></item><item><title>Astro 7.4 adds build timings and parallel prerendering</title><link>https://ia.bokstudio.com/en/news/57fdaeee-astro-7-4-adds-build-timings-and-parallel-prerendering/</link><guid isPermaLink="true">https://ia.bokstudio.com/en/news/57fdaeee-astro-7-4-adds-build-timings-and-parallel-prerendering/</guid><description>The new release adds a --timings flag to see where build time goes, separates build and runtime image services and adds a notFound() handler for custom entrypoints. Parallel prerendering and custom redirect pages arrive as experimental features, useful for large static sites.</description><pubDate>Thu, 15 Oct 2026 00:00:00 GMT</pubDate><source url="https://astro.build/blog/astro-740/">Astro Blog</source><category>Astro</category></item><item><title>Gutenberg 24.2 turns on React 19 by default and moves notes to the editor margin</title><link>https://ia.bokstudio.com/en/news/ea317f62-gutenberg-24-2-turns-on-react-19-by-default-and-moves-notes-to-the-edi/</link><guid isPermaLink="true">https://ia.bokstudio.com/en/news/ea317f62-gutenberg-24-2-turns-on-react-19-by-default-and-moves-notes-to-the-edi/</guid><description>The new Gutenberg plugin release loads React 19 unless an incompatible theme or plugin prevents it, which is worth testing in custom builds. Notes also move to a margin beside the content, Navigation block link search is unified and more blocks gain design controls.</description><pubDate>Wed, 14 Oct 2026 18:03:31 GMT</pubDate><source url="https://make.wordpress.org/core/2026/10/14/whats-new-in-gutenberg-24-2-14-october/">Make WordPress Core</source><category>WordPress</category><category>JavaScript</category></item><item><title>Google releases Gemini 4 Argon, its new frontier model, including via API</title><link>https://ia.bokstudio.com/en/news/ed1be8ea-google-releases-gemini-4-argon-its-new-frontier-model-including-via-ap/</link><guid isPermaLink="true">https://ia.bokstudio.com/en/news/ed1be8ea-google-releases-gemini-4-argon-its-new-frontier-model-including-via-ap/</guid><description>Gemini 4 Argon is Google’s new top model, aimed at multi-step work such as coding and agentic tasks. It is available now through the Google AI Studio API at an introductory $2/$10 per million input/output tokens, doubling after November 30. Google says that inside Antigravity it increased agent code submissions and cut rollbacks.</description><pubDate>Tue, 13 Oct 2026 16:00:00 GMT</pubDate><source url="https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-argon/">Google AI</source><category>Models</category><category>AI for coding</category></item><item><title>Anthropic replaces git with its own system after AI multiplied its merges</title><link>https://ia.bokstudio.com/en/news/b21e647b-anthropic-replaces-git-with-its-own-system-after-ai-multiplied-its-mer/</link><guid isPermaLink="true">https://ia.bokstudio.com/en/news/b21e647b-anthropic-replaces-git-with-its-own-system-after-ai-multiplied-its-mer/</guid><description>Anthropic says its daily merges went from about 2,500 to over 30,000 in a few months because Claude writes much of the code, turning git into the bottleneck. In five months it built Scion, a version control system designed for always-online agents. It is an early look at which workflows stop scaling once agents write the code.</description><pubDate>Tue, 13 Oct 2026 00:00:00 GMT</pubDate><source url="https://www.anthropic.com/engineering/redesigning-source-control">Anthropic Engineering</source><category>AI for coding</category><category>Tools</category></item><item><title>Cloudflare expands Clef with a multimodal model and cuts prices on Workers AI</title><link>https://ia.bokstudio.com/en/news/223b84f7-cloudflare-expands-clef-with-a-multimodal-model-and-cuts-prices-on-wor/</link><guid isPermaLink="true">https://ia.bokstudio.com/en/news/223b84f7-cloudflare-expands-clef-with-a-multimodal-model-and-cuts-prices-on-wor/</guid><description>Clef is Cloudflare’s family of decision models, which return schema-constrained answers instead of free text. The new Clef-omni takes text, images, audio and video in a single call, Clef-flash gets cheaper in exchange for a smaller hosted context, and Clef responds up to twice as fast. Relevant for classifying or moderating content from Workers.</description><pubDate>Fri, 09 Oct 2026 18:27:34 GMT</pubDate><source url="https://blog.cloudflare.com/clef-faster-cheaper-multimodal/">Cloudflare Blog</source><category>Models</category><category>Hosting &amp; deploy</category></item><item><title>The Deno team joins Cloudflare to make self-hosting Workers easier</title><link>https://ia.bokstudio.com/en/news/e056f3d9-the-deno-team-joins-cloudflare-to-make-self-hosting-workers-easier/</link><guid isPermaLink="true">https://ia.bokstudio.com/en/news/e056f3d9-the-deno-team-joins-cloudflare-to-make-self-hosting-workers-easier/</guid><description>Cloudflare is bringing in the Deno team to simplify self-hosting Workers and Durable Objects, so the same building blocks can run outside its network. For developers deploying JavaScript at the edge, it points to less dependence on a single provider.</description><pubDate>Fri, 09 Oct 2026 12:50:00 GMT</pubDate><source url="https://blog.cloudflare.com/deno-joins-cloudflare/">Cloudflare Blog</source><category>Hosting &amp; deploy</category><category>JavaScript</category></item></channel></rss>