agentsclimarketplace

Setup validation

Skill az9713/youtube-transcript-to-skills/.claude/skills/setup-validation

Extract Claude Code skills from AI workflow YouTube transcripts. 10 ready-to-use skills + a meta-skill that automates the extraction process. Built from John Kim's 50 Claude Code tips.

Install
npx -y skills add az9713/youtube-transcript-to-skills --skill setup-validation

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Set up a build/test/lint validation loop for the current project. Detects the tech stack and configures CLAUDE.md with verification commands.

SKILL.md

3.3 KB, 881 tokens by cl100k_base, as published. Nobody here has run it

Setup Validation Loop

Detect the project's tech stack and configure a build/test/lint validation loop in CLAUDE.md.

Usage

/setup-validation

Procedure

Step 1: Detect Tech Stack

Scan the project root for configuration files to identify:

FileIndicates
package.jsonNode.js / JavaScript / TypeScript
tsconfig.jsonTypeScript
Cargo.tomlRust
go.modGo
pyproject.toml or setup.py or requirements.txtPython
GemfileRuby
pom.xml or build.gradleJava/Kotlin
MakefileMake-based build
CMakeLists.txtC/C++
.csproj or *.sln.NET / C#

Step 2: Detect Build System

Based on detected stack, identify the build command:

  • Node.js: Check package.json scripts for build, compile, or tsc
  • Rust: cargo build
  • Go: go build ./...
  • Python: Check for build in pyproject.toml, or python -m build
  • Make: make or make all

If no build step is found, note that as acceptable (some projects don't need one).

Step 3: Detect Test Runner

Identify the test command:

  • Node.js: Check for jest, vitest, mocha, ava in devDependencies. Check package.json scripts for test
  • Rust: cargo test
  • Go: go test ./...
  • Python: Check for pytest, unittest, nose2. Look for pytest.ini or [tool.pytest] in pyproject.toml
  • Ruby: bundle exec rspec or rake test
  • Java: mvn test or gradle test

Also detect single-file test commands (e.g., pytest path/to/test.py) for faster iteration.

Step 4: Detect Linter/Formatter

Identify lint and format commands:

  • Node.js: Check for eslint, prettier, biome in devDependencies. Check scripts for lint, format
  • Rust: cargo clippy (lint), cargo fmt (format)
  • Go: golangci-lint run (lint), gofmt (format)
  • Python: Check for ruff, flake8, black, isort. Prefer ruff check + ruff format if available
  • Ruby: rubocop

Step 5: Present Proposed Validation Loop

Show the user what was detected:

## Detected Validation Loop

Tech stack: [language] + [framework]
Build:      [command or "none needed"]
Test (all): [command]
Test (one): [command] [file]
Lint:       [command]
Format:     [command]

Ask the user to confirm or modify.

Step 6: Update CLAUDE.md

Add or update a "Build & Validation" section in ./CLAUDE.md:

## Build & Validation

- Build: `[command]`
- Test all: `[command]`
- Test single file: `[command] [path]`
- Lint: `[command]`
- Format: `[command]`

Always run the test command after making changes. Run lint before committing.

If CLAUDE.md doesn't exist, create it with this section. If it exists, append this section (or update it if the section already exists).

Step 7: Offer Hook Setup

Ask the user if they'd like to auto-run the formatter after every file edit:

"Would you like to set up a PostToolUse hook to auto-format files after edits? This runs your formatter automatically whenever Claude edits a file."

If yes, suggest running /setup-hooks to configure it.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.