You upgraded your model. Your SKILL.md is still living three releases ago. The skill that was relevant for GPT-5 may add unnecessary work for GPT-6. Same with Claude Code Models. I built Skill Adapter to help old skills adapt without losing the workflow you liked. Original untouched. Changes explained. MIT licensed. github.com/smukh/skill-adapt…

Built Roll Credits, a Claude Mod that gives your coding session movie credits. You direct. Your most-edited file stars. Failed tool calls become plot twists. Runs locally. No extra model calls. MIT licensed. Early access. github.com/smukh/roll-credit…

Most AI coding agents can add an ARIA attribute. Far fewer can prove the original accessibility issue is actually fixed. We built a11y-agent-skills, a portable set of 7 open-source skills for: • Accessibility audits • Source-level fixes • Forms and dialogs • Component reviews • Keyboard navigation • Regression tests Your agent investigates and edits the code. Playwright and axe-core produce deterministic evidence, stable fingerprints, and verification reports. It works across Codex, Claude Code, Cursor, and other Agent Skills-compatible hosts. It also ships as an npm CLI, MCP server, and GitHub Action. Local-first. No telemetry. No API key. Repo in first comment 👇

Everyone is trying to fix AI writing by banning em dashes, “delve” and 50 other AI tells. I think we're fixing the wrong layer. Google, GOV.UK, GitHub, MDN, Apple, Mailchimp etc. have spent YEARS building real editorial systems. So I turned 11 of them into Agent Skills. Instead of: “make this sound human” use: $google-developer-docs $govuk $kubernetes-docs $apple-interface-writing Same model. Different editorial system. Open source ↓

Claude ships watermarks now. I built a skill pack that tells you what actually survives a "watermark remover".

Claude writes. Codex debugs. The circle of AI life.

Ever wondered how one GitHub repo can give you the power of a full engineering team, without the team? It’s called everything-claude-code (159K+ stars, 170+ contributors, Anthropic hackathon winner). Here’s what it actually does and how to use it: • 38 specialized agents that divide the work like a real team: Planner → Architect → Code Reviewer → Security Auditor → Tester → Doc Writer → Build Fixer, and more. • 156 production-ready skills covering real workflows: TDD loops, security scans, refactoring, market research, deployment steps, even investor materials. • Built-in memory & continuous learning It remembers your sessions, extracts useful patterns (“instincts”), turns them into new reusable skills, and gets sharper over time. How to use it (super simple): Install once in Claude Code (or Cursor / Gemini compatible setups): /plugin install everything-claude-code@everything-claude-code Then just type commands like: /ecc:plan "Add user authentication with JWT" → It breaks it down, plans the architecture, writes the code, runs tests, reviews for security, updates docs. Other useful triggers: /tdd, /security-scan, /skill-create from your git history, or /evolve to level up its own capabilities. It turns raw AI coding tools into a structured, self-improving system that handles complex features end-to-end. It's a powerful harness that many devs are now running daily. Check it out and see what it can do for your workflow, link to the repo below 👇👇