Q Claude Skills

A growing collection of Claude skills built around one idea: stop prompting AI and start building systems with it.

Every skill in this repo comes from a real workflow I use or teach. Each one is designed to be installed once and used forever. As I make new content I add new skills here.

New here? Start with three things

If I were starting over with Claude today, /start is the three things I’d build first: a Claude that rewrites its own rules, one prompt that splits into a fleet of agents, and an agent that runs on your machine and texts you. Each links to a working starter.

Want to build your own? docs/make-your-own-skill.md walks you through it.

Want it to work in Codex, Cursor and Copilot too? docs/write-a-skill.md covers the open format and where each agent looks.

Who this is for

Anyone who uses Claude and wants to get more out of it. No coding experience needed.

What’s inside

Thinking

For when you need to understand a problem before you try to solve it.

• Belief Updater — maintains three competing hypotheses and updates them as new evidence comes in
• Blind Spot Finder — finds what is missing from your thinking entirely, not a critique of what you said but a map of what you did not say
• Complexity Mapper — maps all the moving parts of a problem and finds the single lever that creates the most change
• The Five Chains — five four-message sequences for getting Claude past its first answer, every message written out ready to paste
• Five 2026 techniques: Fail, Taste, True, Flip and Bio, five three-prompt chains for finding the failure already underway, your unwritten rules, the assumption a plan rests on, the weaker side of an argument, and who your last month says you are

The thinking skill works in Claude Code, Codex, and Cursor.

Decision Making

For when you need to choose a path and you are too close to see it clearly.

• Parallel Conversations — runs the same idea through three completely different perspectives simultaneously
• Assumption Audit — surfaces every assumption your thinking is based on and stress tests each one
• Time Machine — shows you what your future self would think about this decision in 6 months, 1 year, and 5 years

Research

For when you need to validate an idea or understand a space before committing to it.

• Adversarial Room — builds the strongest possible case against your idea so you find the fatal flaw before it costs you
• Premortem — assumes the project already failed and works backwards to explain why
• Steelman — builds the strongest possible version of any argument so you engage with what is actually true

Building

For when you are ready to execute and need a system not just a plan.

• Weekly Planning — builds a complete weekly plan from your actual context, not generic time blocks
• Agent Setup — designs a full agent workflow covering system prompt, tools, and how to know if it is working
• Research Workflow — turns a research question into a structured workflow with a clear stopping condition
• agent-or-workflow — the sticky note test for deciding if you need an AI agent or just a workflow, with prompts for building both
• check-your-app — three prompts for the three security holes that show up in almost everything built with AI, with the manual checks and what each fix actually is
• claude-watermark — what Claude’s invisible text watermark actually is, what it can and cannot prove, and three things to do about it including disclosure lines you can adapt
• audit-package: before any install or clone, the agent inspects the package, reports five checks in one table, gives a verdict, and waits for you
• write-a-skill: write a skill once and run it in Claude Code, Codex, Cursor and Copilot, with the install path for each

Commands

Templates for slash commands that work across multiple agent frameworks.

• /goal — three-part prompt template for agentic loops (Claude Code, Codex, Hermes). Includes the exact /goal I ran in this repo to build scripts/skill-stats.py, plus shape templates for refactor, research, docs, and content.
• Workflow starter pack — three dynamic workflow prompts for Opus 4.8 in Claude Code: API migration audit, security audit (OWASP Top 10), and README readiness (the workflow from the Opus 4.8 video). Each one ready to paste and save as a slash command.

Self-improving

• /self-improving — a Claude that reads its own results and rewrites its own rules file. One rules file, one signal, a loop that turns evidence into edits. Logs every change and everything is in git, so any edit is one git revert away. The starter includes the self-review skill and an example rules file.

How to install — Claude Desktop and Claude.ai

1. Click the blue Code button on this page
2. Click Download ZIP
3. Unzip the downloaded file
4. Open the unzipped folder and go inside the skills folder
5. Find the folder for the skill you want to install
6. Right-click it and compress it into a ZIP file
7. Open claude.ai
8. Go to Customize, then Skills
9. Click the plus button, then Create skill
10. Select Upload a skill
11. Upload the ZIP file
12. The skill will appear in your Skills list; toggle it on

How to install — Claude Code

npx skills add https://github.com/kju4q/q-claude-skills.git

More coming

Every video I make adds a new skill to this repo. Follow along on Instagram and TikTok.

skill-stats

scripts/skill-stats.py scans every folder under skills/, reads the YAML frontmatter from each SKILL.md, and writes a markdown report to ./skill-stats-report.md with each skill’s name, description, last-modified date, and a flag for any missing required fields (name, description). Python stdlib only — no install step.

python3 scripts/skill-stats.py            # writes ./skill-stats-report.md
python3 scripts/test_skill_stats.py       # runs the test suite against the report

Optional flags: --skills-dir to scan a different tree, --output to write elsewhere.