Writing on AI agents, developer tooling, and software architecture — by Björn Roberg.https://bjro.dev/https://bjro.dev/posts/xai-schema-disclosure-oopsie-or-onpurpose/https://bjro.dev/posts/xai-schema-disclosure-oopsie-or-onpurpose/When xAI open-sourced their For You ranker on 2026-05-15, they redacted every numeric weight, every threshold, every Kafka topic, every config string — but they left every symbol name in source. For a serious competitor, the list of names is the more useful artifact. So either someone messed up, or that was always the plan.Sun, 17 May 2026 00:00:00 GMThttps://bjro.dev/posts/tagging-posts-by-cognitive-mode/https://bjro.dev/posts/tagging-posts-by-cognitive-mode/Adding one optional frontmatter field to my blog produced a taxonomy I couldn't have written without doing the work. The categories come from Bereiter & Scardamalia (1987), and applying them to my own corpus was uncomfortable in useful ways.Tue, 12 May 2026 00:00:00 GMThttps://bjro.dev/posts/memo-deconstructed/https://bjro.dev/posts/memo-deconstructed/MeMo claims near-SOTA prompting via Munger's mental models. Strip the framing: what's left is one trick — ask the LLM to name its reasoning strategy first.Thu, 30 Apr 2026 00:00:00 GMThttps://bjro.dev/posts/gp-deconstructed/https://bjro.dev/posts/gp-deconstructed/Classical GP bundles five independent design choices behind one biology metaphor. Four of them are compute-era assumptions that don't survive contact with frontier LLMs. What's left is a different algorithm.Sat, 18 Apr 2026 00:00:00 GMThttps://bjro.dev/posts/verifier-native-search/https://bjro.dev/posts/verifier-native-search/LLM generation costs roughly 1000× more than verification. The serious program-search systems already exploit this. The agent-scaffold mainstream doesn't — and nobody yet compounds search traces into a domain-specific mutator.Sat, 18 Apr 2026 00:00:00 GMThttps://bjro.dev/posts/claude-code-bash-exit-1-no-output-tmpfs-full/https://bjro.dev/posts/claude-code-bash-exit-1-no-output-tmpfs-full/If Claude Code's Bash tool returns 'Exit code 1' with empty stdout and empty stderr for every command, including `true` and `echo ok`, your /tmp is almost certainly full. Here is the fix.Wed, 15 Apr 2026 10:00:00 GMThttps://bjro.dev/posts/agent-attestation-needs-an-expiry/https://bjro.dev/posts/agent-attestation-needs-an-expiry/The AI-supply-chain cohort is already crowded: Invariant, Snyk Agent Scan, JFrog, Klaimee, a dozen more. They all ship the same artifact, a point-in-time report. That is not what underwriters can price against. The primitive insurance needs is a falsifiable attestation with a TTL that re-verifies on drift. Nobody ships that yet.Wed, 15 Apr 2026 00:00:00 GMThttps://bjro.dev/posts/the-firmbyte-gap/https://bjro.dev/posts/the-firmbyte-gap/The biggest breakthroughs come from connecting unrelated fields — but the mechanism that produces them is broken. Hidalgo's firmbyte concept explains why.Tue, 14 Apr 2026 00:00:00 GMThttps://bjro.dev/posts/ioc-agentic-boundaries/https://bjro.dev/posts/ioc-agentic-boundaries/Applying IoC and dependency injection patterns to agentic service boundaries — and why technical plumbing alone won't prevent control inversion.Tue, 07 Apr 2026 00:00:00 GMThttps://bjro.dev/posts/claude-autohealth/https://bjro.dev/posts/claude-autohealth/Applying signal processing concepts to coding agent actions — from FFT and Markov models to a pragmatic health scoring algorithm for Claude Code hooks.Tue, 24 Mar 2026 00:00:00 GMThttps://bjro.dev/posts/introducing-slog/https://bjro.dev/posts/introducing-slog/slog is a zero-dependency structured logger for Node.js, Deno, Bun, and Cloudflare Workers with pluggable transports and a Hono waitUntil middleware.Sat, 21 Mar 2026 12:00:00 GMThttps://bjro.dev/posts/introducing-cairn/https://bjro.dev/posts/introducing-cairn/cairn is an append-only SQLite event store where immutability is enforced by triggers, not convention. Go, TypeScript, and Rust SDKs share one test spec.Wed, 18 Mar 2026 12:00:00 GMThttps://bjro.dev/posts/new-domain-bjro-dev/https://bjro.dev/posts/new-domain-bjro-dev/The blog has moved to bjro.dev — a short domain that's easier to share and signals a long-term commitment to writing about software and technology.Tue, 17 Mar 2026 12:00:00 GMThttps://bjro.dev/posts/the-npm-problem-nobody-wants-to-work-on/https://bjro.dev/posts/the-npm-problem-nobody-wants-to-work-on/Every Node.js production system depends on a single corporate-controlled npm registry with no real alternative — a structural risk the ecosystem ignores.Thu, 12 Mar 2026 12:00:00 GMThttps://bjro.dev/posts/debian/https://bjro.dev/posts/debian/Debian turns 33 and keeps earning its title as the universal OS — community governance, 59,000 packages, and the foundation of over 1,000 Linux distributions.Sat, 28 Feb 2026 13:00:00 GMThttps://bjro.dev/posts/currying-agents/https://bjro.dev/posts/currying-agents/Apply partial application from functional programming to AI agents — pre-configure agent parameters to create reusable, composable agent building blocks.Wed, 11 Feb 2026 04:00:00 GMThttps://bjro.dev/posts/linear-agentics/https://bjro.dev/posts/linear-agentics/Linear types applied to AI agent safety — a type system that enforces single-use for destructive operations and statically rejects unsafe agent code.Mon, 09 Feb 2026 00:45:00 GMThttps://bjro.dev/posts/markymarkov/https://bjro.dev/posts/markymarkov/MarkyMarkov uses Markov chains to learn patterns from your codebase and provide fast, explainable code guidance that complements LLM agents.Sun, 08 Feb 2026 22:45:00 GMThttps://bjro.dev/posts/building-smarter-ai-agents-with-ideas-from-philosophy/https://bjro.dev/posts/building-smarter-ai-agents-with-ideas-from-philosophy/Philosophically informed AI agent design — modeling belief, knowledge, and epistemic norms — makes agent behavior more reliable and inspectable.Sat, 07 Feb 2026 23:00:00 GMThttps://bjro.dev/posts/markov-chains-and-llms-hybrid-architectures/https://bjro.dev/posts/markov-chains-and-llms-hybrid-architectures/Combine Markov chains with LLMs to build structured, interpretable agents — Markov chains enforce predictable state; LLMs provide semantic flexibility.Sat, 07 Feb 2026 22:45:00 GMThttps://bjro.dev/posts/notes-on-agentic-applications-in-business-processes/https://bjro.dev/posts/notes-on-agentic-applications-in-business-processes/How to pair agentic workflows with Temporal for durability, finite state machines for structure, and MCP servers to scale from pilot to production.Sat, 07 Feb 2026 22:45:00 GMThttps://bjro.dev/posts/context-context-context/https://bjro.dev/posts/context-context-context/Context is the most underrated lever in AI agent design — this post makes the case for MINDMAP-style context objects as a first-class agent primitive.Mon, 02 Feb 2026 13:00:00 GMThttps://bjro.dev/posts/cloud-llms-in-prod/https://bjro.dev/posts/cloud-llms-in-prod/Infra teams quietly feed logs, configs, and schemas to cloud LLMs when troubleshooting — extending the trust boundary into a black box they don’t control.Thu, 29 Jan 2026 13:00:00 GMThttps://bjro.dev/posts/software-development-topics-opinion-1/https://bjro.dev/posts/software-development-topics-opinion-1/Opinionated notes on software development in the age of agentic tooling — where LLMs genuinely change how we work and where the hype outpaces the reality.Sun, 25 Jan 2026 13:00:00 GMThttps://bjro.dev/posts/using-casq-in-ci/https://bjro.dev/posts/using-casq-in-ci/casq brings content-addressable storage to CI pipelines, eliminating redundant builds and test runs by identifying identical inputs with a git-style hash.Sat, 24 Jan 2026 14:00:00 GMThttps://bjro.dev/posts/mise/https://bjro.dev/posts/mise/mise pins tool versions per project — Node, Python, Go, and more — in a single .mise.toml, replacing nvm, pyenv, and Makefiles with one unified tool.Fri, 23 Jan 2026 13:00:00 GMThttps://bjro.dev/posts/whosthere-lan-tool/https://bjro.dev/posts/whosthere-lan-tool/whosthere is a Go-based LAN discovery tool with a modern terminal UI — fast, intuitive network exploration that shows who and what is on your local network.Fri, 23 Jan 2026 13:00:00 GMThttps://bjro.dev/posts/introducing-casq/https://bjro.dev/posts/introducing-casq/casq is a Rust CLI and library for content-addressed file storage — git-style deduplication without the overhead of a full version control system.Thu, 22 Jan 2026 13:00:00 GMThttps://bjro.dev/posts/terminal-multiplexers/https://bjro.dev/posts/terminal-multiplexers/A quick comparison of tmux and zellij as terminal multiplexers — zellij wins on discoverability with visible, mnemonic keybinds right out of the box.Wed, 21 Jan 2026 11:00:00 GMThttps://bjro.dev/posts/locking-down-agents/https://bjro.dev/posts/locking-down-agents/Practical techniques for sandboxing AI agents and limiting blast radius — filesystem restrictions, network controls, and permission scoping in production.Tue, 20 Jan 2026 17:00:00 GMThttps://bjro.dev/posts/making-use-of-the-caps-lock-key/https://bjro.dev/posts/making-use-of-the-caps-lock-key/Remap Caps Lock to Left Control for a more ergonomic keyboard layout — instructions for Debian/Linux via XKBOPTIONS and Windows via SharpKeys.Tue, 20 Jan 2026 13:00:00 GMT