// Skill profile
luongnv89/claude-howto is a how-to-guide skill bundle for Claude Code and any harness honoring the SKILL.md convention. Rather than focusing on a single technical vertical, it packages operator-grade walkthroughs — setup, troubleshooting, workflow patterns, deployment recipes — into discoverable skills that the model can invoke contextually. New on the GitHub trending board 2026-05-16 with +190 stars in 24 hours, it represents a different axis of the skill-bundle taxonomy: instead of domain-vertical (scientific, academic, learning) or generic-directory (mattpocock-style), claude-howto bundles tacit knowledge into a runnable playbook format. Cross-harness compatibility means the same how-tos work across Claude Code, Cursor, and Codex CLI.
luongnv89/claude-howto is a how-to skill bundle for Claude Code and any harness that respects the SKILL.md convention. It hit GitHub trending on 2026-05-16 with +190 stars in 24 hours.
Operator-grade walkthroughs packaged as discoverable Claude skills:
This is a third axis on the 2026 skill-bundle taxonomy. We already have:
claude-howto is the third axis: how-to knowledge as a skill format. Rather than wrapping tools or domain logic, it wraps the tacit operator knowledge that usually lives in wiki pages and Slack threads, and makes it agent-discoverable.
Works with Claude Code, Cursor, Codex CLI, and any harness that honors the .claude/skills/{name}/SKILL.md convention.
See AgentConn’s Skills Directory Race and Skills Go Vertical for the rest of the 2026 skill-bundle landscape.
AI agents that work well with Claude How-To.
Single 70-line CLAUDE.md derived from Karpathy's LLM coding-pitfall observations — #1 GitHub trending for 28 consecutive days.
A complete software development methodology and composable skills framework for coding agents — brainstorm, plan, execute, review, ship.
Agent skill that makes your coding agent think like the laziest senior dev in the room — skip unnecessary code, reuse existing patterns, and write 54% less code while staying 100% safe.