Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
What This Skill Does
Creates Claude Code slash commands: markdown files placed in .claude/commands/ (project) or ~/.claude/commands/ (global) that expand into reusable prompts when invoked with /command-name. Guides through location selection, pattern matching, workflow definition, and file creation with a quality checklist.
Guides selection from four named patterns (Workflow Automation, Iterative Fixing, Agent Delegation, Simple Execution) so commands follow consistent structure rather than being written ad-hoc.
When to use it
- Turning a multi-step code review process into a repeatable /review command
- Building a /fix-lint command that iterates until CI passes
- Packaging a deployment checklist as a single invokable command
- Creating a project-specific /scaffold command for new feature setup
- Automating a recurring git workflow like daily standup summaries
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: Create a command that runs tests, parses failures, and fixes them until they pass
- 1Run git rev-parse to detect repo context and select .claude/commands/ as the target directory
- 2Present four patterns and identify Iterative Fixing as the match
- 3Ask for command name, argument requirements, and specific test runner command
- 4Generate command file using imperative-form steps with explicit tool calls and success criteria
- 5Write the file and report location, usage syntax, and how to invoke it
.claude/commands/fix-tests.md created, invokable as /fix-tests with documented iteration logic and exit conditions