// 12 files
Anthropic's new Mythos-class model, Claude Fable 5, is its strongest coding and agentic model yet — cheaper than the preview, with a new safeguard layer that can fall back to Opus 4.8.
AI did not kill engineering. It killed the illusion that typing code was the hard part. In my Claude Code + Go + oCMS workflow, the human job moved upstream: intent, constraints, architecture, verification, and context discipline.
A 54-commit build diary: Claude Desktop drafted the spec, Claude Code executed it, and the result is a production-grade single-binary Go service scheduled by systemd.
Karpathy-style LLM Wiki over oCMS: 106 Markdown pages compiled by Claude Code in one evening, ten documentation contradictions surfaced, wikilint enforcing invariants, CLAUDE.md making the wiki consultable for agents.
Andrej Karpathy — former Tesla AI director, OpenAI co-founder, and one of the most influential voices in AI — has been quietly sketching a map of where software development is heading. Across...
The Model Context Protocol is everywhere. Claude Code, ChatGPT, VS Code Copilot, Cursor — they all speak MCP now. But if you search for "how to build an MCP server," you'll drown in Python and...
Claude Code 2.1.97 just landed with 46 CLI changes, 1 flag change, and 2 system prompt changes. This is one of the heaviest patch releases in recent memory — and most of the weight is in fixes. If...
Two Claude Code releases dropped today within seven hours of each other. Version 2.1.94 is a substantial update — 25 CLI changes and 2 system prompt changes — while 2.1.96 is a targeted hotfix...
How a pen test finding turned into a shipped feature — planned, coded, tested, and pushed to production from a phone while walking a dog along the Rhône, using Claude Code in the browser.
Claude Code ignores plan mode and edits files anyway? Here's why it happens and 5 battle-tested techniques — from hooks to prompt architecture — to enforce mode discipline.