An engineering journal built in public: AI agent systems, Traffic OS, and what breaks in production.https://enriquerubio.dev/enhttps://enriquerubio.dev/en/blog/where-private-data-lives-local-ai-setup/https://enriquerubio.dev/en/blog/where-private-data-lives-local-ai-setup/A walk through the channels private data can leak through in a local AI setup (Git, bind mounts, builds, logs, worktrees) and how one directory was picked so that none of them reaches it.Tue, 20 Oct 2026 00:00:00 GMTarchitectureprivacydockergitlocal-firstarchitectureEnrique Rubiohttps://enriquerubio.dev/en/blog/lufs-correct-is-not-good-audio-two-gates/https://enriquerubio.dev/en/blog/lufs-correct-is-not-good-audio-two-gates/An audio bed hit its loudness target and a person still judged it too synthetic. How the code keeps the machine verdict and the human verdict apart.Mon, 19 Oct 2026 00:00:00 GMTbuildingaudiolufsquality-gatesffmpeghuman-in-the-loopEnrique Rubiohttps://enriquerubio.dev/en/blog/claude-code-fixes-unattended-agent/https://enriquerubio.dev/en/blog/claude-code-fixes-unattended-agent/The latest Claude Code release notes include small changes that matter if you leave it working unwatched: worktrees, binary files, search errors and usage limits.Fri, 16 Oct 2026 00:00:00 GMTai-newsclaude-codeagentesrelease-notesautomatizacionEnrique Rubiohttps://enriquerubio.dev/en/blog/gemini-4-argon-ai-gateway-thinking-levels/https://enriquerubio.dev/en/blog/gemini-4-argon-ai-gateway-thinking-levels/Vercel added Gemini 4 Argon to AI Gateway. What it changes for builders: the model ID, level-based reasoning control, and how pricing is passed through.Wed, 14 Oct 2026 00:00:00 GMTai-newsgeminiai-gatewayvercelllmEnrique Rubiohttps://enriquerubio.dev/en/blog/docker-rm-f-anonymous-volumes-postgres-test-leak/https://enriquerubio.dev/en/blog/docker-rm-f-anonymous-volumes-postgres-test-leak/My test fixtures removed ephemeral Postgres containers with docker rm -f and left an anonymous volume behind per container. What the code shows, and the regression test with a negative control.Mon, 12 Oct 2026 00:00:00 GMTbuildingdockerpostgrespytesttest-fixturesanonymous-volumesEnrique Rubiohttps://enriquerubio.dev/en/blog/ai-system-passed-2-of-3-tests-to-publish-alone/https://enriquerubio.dev/en/blog/ai-system-passed-2-of-3-tests-to-publish-alone/My blog pipeline passed 2 of 3 autonomous-publishing runs. The third was refused, so it stays on human approval. Why I didn't loosen the approval gates.Sat, 10 Oct 2026 00:00:00 GMTarchitectureAI agentsapproval gateshuman-in-the-loopAI agent safetyautonomous publishingbuild-in-publicEnrique Rubiohttps://enriquerubio.dev/en/blog/linkedin-timeout-unconfirmed-not-retry/https://enriquerubio.dev/en/blog/linkedin-timeout-unconfirmed-not-retry/How Jarvis handles a timeout when posting to LinkedIn: a state of its own for 'I do not know', and a human as the only way to force a retry.Sat, 10 Oct 2026 00:00:00 GMTbuildinglinkedinidempotencypublishingstate-machinebuild-in-publicEnrique Rubiohttps://enriquerubio.dev/en/blog/my-own-video-editing-pipeline-1-8-gb-raw-to-youtube/https://enriquerubio.dev/en/blog/my-own-video-editing-pipeline-1-8-gb-raw-to-youtube/How my pipeline turned ~1.8 GB of raw video into an 8-minute 1080p YouTube video: upload, audio, motion, metadata, chapters, thumbnails and human review.Sat, 10 Oct 2026 00:00:00 GMTbuildingvideo editing pipelinelong-form videoYouTube workflowAI agentsbuild-in-publicEnrique Rubiohttps://enriquerubio.dev/en/blog/openai-dots-always-on-agents-vs-metas-free-muse/https://enriquerubio.dev/en/blog/openai-dots-always-on-agents-vs-metas-free-muse/OpenAI Dots: eligible Pro and Business Premium users get one agent at no extra charge; extra workload has no price yet. Meta's Muse is free for most uses.Fri, 02 Oct 2026 00:00:00 GMTai-newsopenaimetaagentes-iadotsmuseEnrique Rubiohttps://enriquerubio.dev/en/blog/gemini-4-argon-cyber-defenders-first/https://enriquerubio.dev/en/blog/gemini-4-argon-cyber-defenders-first/Gemini 4 Argon starts with trusted defenders in Google's Fairwind Program. Launch price: $2/$10 per million tokens, then $4/$20. Benchmarks are self-reported.Thu, 01 Oct 2026 00:00:00 GMTai-newsgooglegeminiciberseguridadfairwindia-generativaEnrique Rubiohttps://enriquerubio.dev/en/blog/openai-cancels-gpt-6-1-astra-authorization-honesty/https://enriquerubio.dev/en/blog/openai-cancels-gpt-6-1-astra-authorization-honesty/OpenAI drops GPT-6.1 Astra because it missed the bar on scope, authorization and honesty with the user. What failed and what it means for agent builders.Wed, 30 Sep 2026 00:00:00 GMTai-newsopenaigpt-6-1seguridad-iaagentesalineamientoEnrique Rubiohttps://enriquerubio.dev/en/blog/mentalhealthbench-openai-gpt-6-astra-self-graded/https://enriquerubio.dev/en/blog/mentalhealthbench-openai-gpt-6-astra-self-graded/OpenAI's new mental-health benchmark puts GPT-6 Astra on top, ahead of Anthropic's Claude Opus 5.5 and Google's Gemini 2.5 Pro -- graded by another OpenAI model.Sun, 27 Sep 2026 00:00:00 GMTai-newsopenaigpt-6-astramentalhealthbenchsalud-mental-iaevaluacion-de-iaEnrique Rubiohttps://enriquerubio.dev/en/blog/gemini-live-avatar-google-model-not-new/https://enriquerubio.dev/en/blog/gemini-live-avatar-google-model-not-new/Google's Live Avatar streams a lip-synced face onto Gemini's live conversations in 97 languages -- but its own safety report says the model underneath isn't new.Fri, 25 Sep 2026 00:00:00 GMTai-newsgooglegeminigemini-live-avataravatares-iaseguridad-iaEnrique Rubiohttps://enriquerubio.dev/en/blog/gpt-6-astra-parallel-monitorability-openai/https://enriquerubio.dev/en/blog/gpt-6-astra-parallel-monitorability-openai/OpenAI is showcasing Parallel as GPT-6 Astra's success story: half the time, half the cost. The same week, the model's own safety report admits a substantial drop in internal transparency -- and OpenAI's own researchers say they're worried.Thu, 24 Sep 2026 00:00:00 GMTai-newsopenaigpt-6-astraparallelagentes-iaseguridad-iaEnrique Rubiohttps://enriquerubio.dev/en/blog/ringg-openai-four-models-automated-support/https://enriquerubio.dev/en/blog/ringg-openai-four-models-automated-support/OpenAI published a customer story about Ringg, a voice-agent platform that automates customer support in India: up to 65% of routine inquiries get resolved without a human, by routing work to four different models based on the task. The routing pattern is more interesting than the headline number.Wed, 23 Sep 2026 00:00:00 GMTai-newsAI agentsOpenAIautomationvoice AIEnrique Rubiohttps://enriquerubio.dev/en/blog/traffic-os-first-production-failure/https://enriquerubio.dev/en/blog/traffic-os-first-production-failure/Traffic OS's first real public canary was one command away from attributing a private-validation video's metrics to a different, public video. Here's what broke, how I caught it, and how I fixed it.Wed, 23 Sep 2026 00:00:00 GMTbuildingTraffic OSAI agentsYouTube APIproduction incidentsEnrique Rubiohttps://enriquerubio.dev/en/blog/the-silent-bug-that-stalled-my-content-pipeline/https://enriquerubio.dev/en/blog/the-silent-bug-that-stalled-my-content-pipeline/My agent pipeline still looked technically healthy — the crons ran, nothing errored — but its state machine was stuck: nothing ever moved from 'planned' to 'recorded'. Here's what was missing, why it took me so long to see it, and how I fixed it.Tue, 22 Sep 2026 00:00:00 GMTarchitectureJarvisAI agentscontent pipelinesproduction incidentsPostgreSQLEnrique Rubio