A Hacker News user known as phasE89 has published a free, automatically-updated reference guide for Claude Code at cc.storyfox.cz, covering the full surface area of Anthropic's CLI coding assistant. The cheat sheet spans keyboard shortcuts, over 30 slash commands, CLI flags, MCP server configuration, the skills and agents system, memory management, and key environment variables — with the current version of Claude Code (v2.1.81 as of March 23, 2026) displayed at the top alongside a dismissable changelog panel. The resource is a single HTML file, requires no signup, and auto-detects Mac versus Windows to surface the correct keyboard shortcut variants.
What distinguishes the project technically is its pipeline architecture: a daily cron job reads the official Claude Code GitHub changelog, identifies new features relative to the previous run, and regenerates the HTML with "NEW" badges on recently added entries. The March 23 update highlights include a --bare flag for minimal headless mode, a --channels flag for permission relay and MCP push messages (currently in preview), effort frontmatter support for skills and slash commands, the renaming of /fork to /branch with a legacy alias, and SendMessage auto-resumption of stopped agents. The cheat sheet itself was built using Claude Code — phasE89 had Claude research its own documentation and changelog, then generate the printable A4 landscape layout.
The Hacker News discussion surfaced two points worth tracking. Commenters noted the cheat sheet omits the --dangerously-skip-permissions flag, a meaningful gap for power users running automated or trusted workflows. The thread also produced a clear consensus that <a href="/news/2026-03-14-emacs-vim-ai-terminal-native-advantage">Claude Code is substantially ahead of OpenAI Codex on CLI capabilities</a> — framing the cheat sheet not just as a convenience resource but as indirect evidence of how feature-dense and complex Claude Code has become. The presence of CLAUDE_CODE_DISABLE_CRON in the env vars table is one concrete example: Anthropic has shipped a dedicated kill-switch for cron scheduling directly into the environment variable surface.
The --dangerously-skip-permissions omission is the one obvious gap phasE89 hasn't addressed yet. Anthropic keeps the flag out of its prominent documentation, which raises a practical question: whether a changelog-driven pipeline will ever surface it automatically, or whether community annotation will have to fill that hole by hand.