PlugWiki

wizard Skill for Claude Code: Generate a Step-by-Step Setup Script

Writes an interactive bash script that walks you through steps only a person can do, like getting API keys, setting CI secrets or running a cutover.

Installs
587.5K
on skills.sh
Stars
309.2K
on GitHub
Files
2
besides SKILL.md
Updated
Nov 2026
last repo commit

wizard is for work the agent cannot do for you: provisioning infrastructure, creating credentials, clicking through a third-party dashboard or running a one-off migration. The agent reads your repo (.env files, README, compose files, workflows) to list every step and value, shows you the stages, then fills in a bundled bash template.

The script opens each URL, tells you what to click and copy, hides secret entry, writes values to .env and can set GitHub secrets and variables with the gh CLI. It is meant to be thrown away after one run unless you want it in the repo.

Good for

  • Developers setting up credentials, CI secrets or cloud accounts
  • Teams who keep re-explaining the same setup to an AI
  • Anyone running a one-off migration with irreversible steps

Install wizard

Claude Code

  1. 1

    Copy this prompt.

    Install the wizard skill in Claude Code. Run this command:
    
    npx skills add https://github.com/mattpocock/skills --skill wizard
    
    Then check that it loaded and tell me how to use it.
  2. 2

    Paste it into Claude Code. It shows the change and asks before running it.

Requires Node.js.

From its listing on skills.sh. Source.

What's in the skill

When it loads

Generate an interactive bash wizard that walks a human through steps only they can perform. Use when provisioning infrastructure, setting up credentials or CI secrets, walking an unfamiliar third-party dashboard, or running a one-off migration or cutover. Don't invoke this for steps the agent can perform itself.

Sections in SKILL.md

  1. Process
  2. 1. Scope the procedure
  3. 2. Map each stage's journey
  4. 3. Author the wizard
  5. 4. Verify and hand off

Key facts

Installs on skills.sh
587,475 (Nov 2, 2026)
License
MIT
Last repo commit
Nov 1, 2026
Files next to SKILL.md
2 files

What it includes

  • Reads your repo and lists every manual step and captured value, then confirms with you

    Read · 1. Scope the procedure

  • Writes the exact path a person follows: URL, click, copy, target variable

    2. Map each stage

  • Copies template.sh and fills in one stage per step

    Write · 3. Author the wizard

  • Runs bash -n and shellcheck if present, then tells you how to run it

    Runs code · 4. Verify and hand off

  • Helpers: stage, open_url, ask_secret, write_env, set_secret, set_var, confirm

    Running the finished script opens browser pages, writes .env and can set GitHub secrets. The agent is told not to run it itself.

    Runs code · template.sh

From the official docs

Permissions

Medium risk

What it can do

  • Read your data
  • Create and change things
  • Delete things
  • Read-only mode

Watch out for

  • The script it writes can set .env values and GitHub secrets

    Read the script before you run it. Check each set_secret name against your CI workflow. github.com

  • The generated script runs shell commands

    The agent only checks it statically. You run it, so review it first and delete it after one-off jobs. github.com

  • Secrets end up in .env

    Keep .env out of git and out of shared folders. github.com

Technical details
Talks toOpens URLs in your browser. GitHub, through gh, if the script sets secrets
Scopes
  • Your repo's .env files, README and workflows (read)

  • Values you type into the finished script

Questions

What does wizard do?

It generates a bash script that guides a person through manual setup. The agent scopes the steps and writes the script, and you run it yourself.

How do I install wizard?

Use the install box at the top. The skill is one SKILL.md plus template.sh, and it links to no other skill.

Does the agent run the script?

No. The SKILL.md says not to run it end to end, because it opens browsers and waits for human input. The agent only checks it with bash -n and shellcheck and by reading it.

Does it handle my secrets?

The template uses hidden entry for secrets and can write them to .env and set them as GitHub secrets. That is the point of it, so run it on your own machine, and keep .env out of git.

Do I need the GitHub CLI?

Only for the stages that set GitHub secrets or variables. The template's helpers for that use gh.

Is it safe?

The script it writes can modify .env and your GitHub repo settings when you run it. Read it before you run it, and delete it afterwards unless it should live in the repo.

Information

updated Nov 2, 2026

Folder
skills/engineering/wizard

mattpocock/skillsskills.sh

Sources

Docs, issue trackers and discussions this page is built from.

Show all 3
  1. wizard SKILL.md · github.com
  2. wizard on skills.sh · skills.sh
  3. mattpocock/skills README · github.com

Updated Nov 2, 2026.