<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Hugo Nogueira</title><description>Hugo Nogueira is CPTO at Complyance. He writes about running AI agents in production, with a focus on security and reliability, and publishes open research on how agents fail.</description><link>https://hugo.im/</link><item><title>Machines Read, Humans Sign: Code Review in One Year</title><link>https://hugo.im/posts/code-review-in-one-year/</link><guid isPermaLink="true">https://hugo.im/posts/code-review-in-one-year/</guid><description>What building our own review agent taught us, with our numbers, what Anthropic and one open-source pull request show about where review is going, and eight bets on how teams will review code by October 2027.</description><pubDate>Tue, 20 Oct 2026 22:00:00 GMT</pubDate></item><item><title>Which Keys Is Your Agent Holding?</title><link>https://hugo.im/posts/heldkeys/</link><guid isPermaLink="true">https://hugo.im/posts/heldkeys/</guid><description>A coding agent running in your terminal can use every credential that terminal can reach. heldkeys is an open-source CLI that lists those credentials on your laptop, inside a sandbox or in a CI job, and tells you what to fix first.</description><pubDate>Fri, 16 Oct 2026 00:10:00 GMT</pubDate></item><item><title>A Capability, Not a Credential: How Our AI Agents Call Customer APIs Without Ever Holding a Key</title><link>https://hugo.im/posts/capability-not-credential/</link><guid isPermaLink="true">https://hugo.im/posts/capability-not-credential/</guid><description>Our AI agents write and run code against customers&apos; GitHub, AWS and other systems from a sandbox with internet access, and nothing inside that sandbox can authenticate to any of them. The architecture behind it, the designs we set aside, and where its boundaries are.</description><pubDate>Fri, 09 Oct 2026 08:00:00 GMT</pubDate></item><item><title>What 639,000 Execution Steps Taught Me About How AI Agents Really Fail</title><link>https://hugo.im/posts/mast-production-agent-failures/</link><guid isPermaLink="true">https://hugo.im/posts/mast-production-agent-failures/</guid><description>I applied the MAST failure taxonomy to 639,000 execution steps from AI agents running in production for five months. My first headline finding turned out to be an infrastructure bug masquerading as agent behavior. This post is about what agents actually fail at in production, and the discipline it takes to not fool yourself with production data.</description><pubDate>Sat, 30 May 2026 18:00:00 GMT</pubDate></item><item><title>LHC v0.2: A Benchmark for Long-Horizon Agent Coherence (and the Methodology That Got It Honest)</title><link>https://hugo.im/posts/lhc-v02-long-horizon-coherence-benchmark/</link><guid isPermaLink="true">https://hugo.im/posts/lhc-v02-long-horizon-coherence-benchmark/</guid><description>I just published LHC v0.2, an open benchmark for long-horizon coherence in 8B-class agent models, plus a deterministic parser baseline that puts a useful floor on what fine-tuning is worth for structured-state tasks. This post explains what they&apos;re for, how to use them, and the methodology arc that produced them across five rounds of external review.</description><pubDate>Sun, 10 May 2026 20:00:00 GMT</pubDate></item><item><title>We&apos;re Mistaking the Bootstrap Phase for the Future of AI Agents</title><link>https://hugo.im/posts/bootstrap-phase-future-ai-agents/</link><guid isPermaLink="true">https://hugo.im/posts/bootstrap-phase-future-ai-agents/</guid><description>The self-hosted AI agent movement is real and important. But we are confusing a bootstrap phase with a destination architecture. The long-term future of agents will be defined by platforms that make them reliable, governable, and operationally boring.</description><pubDate>Wed, 01 Apr 2026 08:00:00 GMT</pubDate></item><item><title>Execution Is Cheap. Judgment Isn&apos;t: AI Agents and the Collapse of the CTO/CPO Divide</title><link>https://hugo.im/posts/execution-is-cheap-judgment-isnt/</link><guid isPermaLink="true">https://hugo.im/posts/execution-is-cheap-judgment-isnt/</guid><description>When execution becomes abundant through AI agents, judgment becomes the bottleneck. The traditional separation between technical and product leadership breaks down, creating space for the CPTO role.</description><pubDate>Sat, 21 Feb 2026 10:00:00 GMT</pubDate></item><item><title>The OWASP Top 10 for AI Agents - Security in the Age of Autonomy</title><link>https://hugo.im/posts/owasp-top10-ai-agents/</link><guid isPermaLink="true">https://hugo.im/posts/owasp-top10-ai-agents/</guid><description>OWASP just published their first Top 10 for Agentic Applications. Here&apos;s what every agent builder needs to know about the new attack surfaces, why traditional security fails, and how the orchestration layer becomes the new security boundary.</description><pubDate>Mon, 09 Feb 2026 08:00:00 GMT</pubDate></item><item><title>From OpenClaw&apos;s Chaos to OpenAI&apos;s Frontier: The Agent Infrastructure Reckoning</title><link>https://hugo.im/posts/agent-security-paradox/</link><guid isPermaLink="true">https://hugo.im/posts/agent-security-paradox/</guid><description>OpenClaw went from viral sensation to security nightmare. Today, OpenAI launched Frontier. Two sides of the same story: autonomous agents are going mainstream, and the infrastructure isn&apos;t ready.</description><pubDate>Thu, 05 Feb 2026 10:00:00 GMT</pubDate></item><item><title>The 100th Tool Call Problem: Why Most AI Agents Fail in Production</title><link>https://hugo.im/posts/100th-tool-call-problem/</link><guid isPermaLink="true">https://hugo.im/posts/100th-tool-call-problem/</guid><description>Demos run for minutes. Production runs for hours. Here&apos;s what breaks, and the durability patterns that actually work when your agent needs to make 200+ tool calls without losing its mind.</description><pubDate>Tue, 27 Jan 2026 08:00:00 GMT</pubDate></item><item><title>Specs Are Not Free: Why AI Won&apos;t Replace Programming (It Will Transform It)</title><link>https://hugo.im/posts/specs-are-not-free/</link><guid isPermaLink="true">https://hugo.im/posts/specs-are-not-free/</guid><description>A response to the viral &quot;spec as code&quot; thesis. We don&apos;t stop programming. We change what programming means. And that change requires more engineering fundamentals, not fewer.</description><pubDate>Wed, 14 Jan 2026 10:00:00 GMT</pubDate></item><item><title>The Agent Harness: Why 2026 is About Infrastructure, Not Intelligence</title><link>https://hugo.im/posts/agent-harness-infrastructure/</link><guid isPermaLink="true">https://hugo.im/posts/agent-harness-infrastructure/</guid><description>Intelligence without infrastructure is just a demo. Here&apos;s why the Agent Harness thesis matters and what I&apos;ve learned building autonomous agents that actually work.</description><pubDate>Wed, 07 Jan 2026 22:00:00 GMT</pubDate></item><item><title>2025: The Year AI Agents Got Real</title><link>https://hugo.im/posts/ai-agents-year-review-2025/</link><guid isPermaLink="true">https://hugo.im/posts/ai-agents-year-review-2025/</guid><description>A practitioner&apos;s look back at what happened in AI agents in 2025, from DeepSeek shaking the markets to GPT-5, Gemini 3, and the rise of agentic AI.</description><pubDate>Sat, 03 Jan 2026 08:00:00 GMT</pubDate></item><item><title>The Memory Problem in AI Agents</title><link>https://hugo.im/posts/memory-problem-ai-agents/</link><guid isPermaLink="true">https://hugo.im/posts/memory-problem-ai-agents/</guid><description>Why memory architecture is the hidden bottleneck in AI agent systems, and the patterns that actually work in production.</description><pubDate>Sat, 27 Dec 2025 14:00:00 GMT</pubDate></item><item><title>Where&apos;s the shovelware? Right here. Why AI coding works (if you know how to use it)</title><link>https://hugo.im/posts/ai-coding-productivity-response/</link><guid isPermaLink="true">https://hugo.im/posts/ai-coding-productivity-response/</guid><description>What changed when I used AI coding tools seriously for four months, with my GitHub data, the workflows behind it, and a response to the skeptics based on 146,000 lines shipped.</description><pubDate>Fri, 05 Sep 2025 14:00:00 GMT</pubDate></item><item><title>Why I stopped estimating: a data-driven case against software predictions</title><link>https://hugo.im/posts/why-i-stopped-estimating/</link><guid isPermaLink="true">https://hugo.im/posts/why-i-stopped-estimating/</guid><description>A data-driven analysis of why software estimation doesn&apos;t work and what to do instead. Based on research, industry reports, and real-world experience with the no-estimates movement.</description><pubDate>Sat, 28 Jun 2025 15:30:00 GMT</pubDate></item><item><title>How to strike the right balance between not rushing to show value and exceeding expectations in early-stage startups</title><link>https://hugo.im/posts/balancing-the-cofounder-mindset/</link><guid isPermaLink="true">https://hugo.im/posts/balancing-the-cofounder-mindset/</guid><description>Explore the significance of balancing value and a co-founder mindset in early-stage startups. Learn how to show value while respecting team dynamics and collaborating towards common goals.</description><pubDate>Mon, 22 Jul 2024 09:15:00 GMT</pubDate></item><item><title>Creating an efficient and healthy interview process for engineers: insights from a decade of leading engineering teams</title><link>https://hugo.im/posts/healthy-and-inclusive-engineering-interview-process/</link><guid isPermaLink="true">https://hugo.im/posts/healthy-and-inclusive-engineering-interview-process/</guid><description>Create an efficient interview process for engineers. Learn practical tips for assessing skills, fit, and diversity, attracting top talent, and mitigating biases.</description><pubDate>Fri, 15 Mar 2024 14:20:00 GMT</pubDate></item><item><title>The importance of zero trust architecture for enterprises with remote or hybrid work models</title><link>https://hugo.im/posts/the-importance-of-zero-trust-architecture-for-enterprises-with-remote-or-hybrid-work-models/</link><guid isPermaLink="true">https://hugo.im/posts/the-importance-of-zero-trust-architecture-for-enterprises-with-remote-or-hybrid-work-models/</guid><description>Learn about Zero Trust Architecture for remote and hybrid work. This approach to cybersecurity provides stronger security and lower risk for businesses.</description><pubDate>Tue, 09 May 2023 10:00:00 GMT</pubDate></item><item><title>The world is not that bad! Book review: Factfulness</title><link>https://hugo.im/posts/book-review-factfulness/</link><guid isPermaLink="true">https://hugo.im/posts/book-review-factfulness/</guid><description>Explore ‘Factfulness’ by Hans Rosling and learn how the world is improving. Discover how our instincts can distort reality and read about global progress.</description><pubDate>Sun, 10 Jan 2021 08:30:00 GMT</pubDate></item><item><title>A brief look at hey.com stack</title><link>https://hugo.im/posts/a-brief-look-at-heycom-stack/</link><guid isPermaLink="true">https://hugo.im/posts/a-brief-look-at-heycom-stack/</guid><description>An exploration of the technology stack and infrastructure behind Basecamp&apos;s Hey email service, including Ruby on Rails, Kubernetes, and their development workflow.</description><pubDate>Tue, 30 Jun 2020 00:00:00 GMT</pubDate></item><item><title>Avoiding burnout as a software engineer</title><link>https://hugo.im/posts/avoiding-burnout-as-a-software-engineer/</link><guid isPermaLink="true">https://hugo.im/posts/avoiding-burnout-as-a-software-engineer/</guid><description>Learn how to recognize, prevent and fight burnout as a software engineer, with practical tips for both individuals and managers to create healthier work environments.</description><pubDate>Sat, 14 Dec 2019 00:00:00 GMT</pubDate></item><item><title>What diversity should not be about, especially in software teams</title><link>https://hugo.im/posts/what-diversity-should-not-be-about-especially-in-software-teams/</link><guid isPermaLink="true">https://hugo.im/posts/what-diversity-should-not-be-about-especially-in-software-teams/</guid><description>A thoughtful perspective on diversity in software teams, exploring what true diversity means and what practices should be avoided when promoting inclusivity.</description><pubDate>Fri, 25 Jan 2019 00:00:00 GMT</pubDate></item><item><title>My first exit: what I&apos;ve learned in 8 years leading meuingresso.com</title><link>https://hugo.im/posts/my-first-exit/</link><guid isPermaLink="true">https://hugo.im/posts/my-first-exit/</guid><description>The story of my first startup exit after 8 years building and leading meuingresso.com, including lessons learned about execution, trust, and the entrepreneurial journey.</description><pubDate>Wed, 16 Jan 2019 00:00:00 GMT</pubDate></item><item><title>How to develop reusable components with Babel and RollupJS</title><link>https://hugo.im/posts/how-to-develop-reusable-components-with-babel-and-rollupjs/</link><guid isPermaLink="true">https://hugo.im/posts/how-to-develop-reusable-components-with-babel-and-rollupjs/</guid><description>A comprehensive guide on creating and publishing reusable JavaScript components using Rollup.js and Babel, including the evolution of JavaScript modules.</description><pubDate>Wed, 26 Sep 2018 00:00:00 GMT</pubDate></item><item><title>var, let or const?</title><link>https://hugo.im/posts/var-let-or-const/</link><guid isPermaLink="true">https://hugo.im/posts/var-let-or-const/</guid><description>Understanding the differences between var, let, and const in JavaScript ES6, including scope behaviors and when to use each variable declaration.</description><pubDate>Mon, 06 Aug 2018 00:00:00 GMT</pubDate></item><item><title>My personal mission statement</title><link>https://hugo.im/posts/my-personal-mission-statement/</link><guid isPermaLink="true">https://hugo.im/posts/my-personal-mission-statement/</guid><description>My personal mission statement defining how I approach life, work, relationships, and personal growth, inspired by Stephen Covey&apos;s principles.</description><pubDate>Wed, 01 Aug 2018 00:00:00 GMT</pubDate></item></channel></rss>