create-cli_skill

Overview

This skill designs the command-line interface parameters and user experience for a tool before implementation. It produces a compact, implementable CLI specification covering command tree, flags, I/O, error codes, config precedence, safety and example invocations.

How this skill works

I ask a small set of clarifying questions (command name, primary user, input/output contracts, interactivity and config model) and apply a default rubric based on CLI best practices. Then I generate a concise spec: usage synopsis, args/flags table, subcommand semantics, I/O rules, exit codes, config precedence, safety behaviors, shell completion notes, and 5–10 example invocations.

When to use it

• Designing a new CLI surface before writing code.
• Refactoring or standardizing an existing CLI for consistency.
• Specifying machine- and human-friendly output formats.
• Defining safety and non-interactive behaviors for automation.
• Auditing CLI discoverability and completion/install story.

Best practices

• Ask minimal clarifying questions; use sensible defaults if unsure.
• Keep primary data on stdout and diagnostics on stderr; add --json and --plain.
• Make prompts conditional on TTY; provide --no-input and --force for automation.
• Follow clear precedence for flags > env > project config > user config > system.
• Define explicit exit codes for parse/validation, generic failures, and common error modes.
• Document example invocations including piped stdin and config-file scenarios.

Example use cases

• Design a new deploy CLI: name, subcommands, flags, dry-run, and confirmation rules.
• Refactor a backup tool to add --json output and consistent exit codes for automation.
• Specify config precedence and environment variables for a cross-platform single-binary tool.
• Create a help/usage tree and completion install instructions for discoverability.
• Add safe defaults: --dry-run, --no-input, and explicit --confirm token for destructive ops.