Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Generate formatted help text with examples, descriptions, sections, and consistent styling for CLI applications.
.claude/skills/a5c-ai-help-text-formatter/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | 2% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 14% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 22% | 0% |
| case-14 | ✗→✓ | ▲ Improved | 54% | 0% |
| case-17 | ✗→✓ | ▲ Improved | 80% | 0% |
Generate formatted, user-friendly help text for CLI applications.
Invoke this skill when you need to:
| Parameter | Type | Required | Description | |-----------|------|----------|-------------| | language | string | Yes | Target language (typescript, python, go) | | commands | array | Yes | Command definitions with help text | | styling | object | No | Color and formatting options |
json{ "commands": [ { "name": "deploy", "description": "Deploy application to target environment", "longDescription": "Deploy the application with optional configuration...", "arguments": [ { "name": "service", "description": "Service to deploy" } ], "options": [ { "flags": "-e, --env", "description": "Target environment" } ], "examples": [ { "command": "deploy api -e production", "description": "Deploy API to production" } ] } ] }
typescriptimport chalk from 'chalk'; interface HelpSection { title: string; content: string | string[]; } export function formatHelp(command: CommandDefinition): string { const sections: HelpSection[] = []; // Description sections.push({ title: '', content: command.description, }); // Usage sections.push({ title: 'Usage', content: ` $ ${command.name} ${formatUsage(command)}`, }); // Arguments if (command.arguments?.length) { sections.push({ title: 'Arguments', content: command.arguments.map(arg => ` ${chalk.cyan(arg.name.padEnd(20))} ${arg.description}` ), }); } // Options if (command.options?.length) { sections.push({ title: 'Options', content: command.options.map(opt => ` ${chalk.cyan(opt.flags.padEnd(20))} ${opt.description}` ), }); } // Examples if (command.examples?.length) { sections.push({ title: 'Examples', content: command.examples.map(ex => ` ${chalk.dim('$')} ${ex.command}\n ${chalk.dim(ex.description)}` ), }); } return renderSections(sections); } function renderSections(sections: HelpSection[]): string { return sections.map(section => { const title = section.title ? chalk.bold.underline(section.title) + '\n\n' : ''; const content = Array.isArray(section.content) ? section.content.join('\n') : section.content; return title + content; }).join('\n\n'); }
pythonfrom typing import List, Optional import textwrap class HelpFormatter: def __init__(self, width: int = 80, indent: int = 2): self.width = width self.indent = ' ' * indent def format_command(self, command: dict) -> str: sections = [] # Description sections.append(command['description']) # Usage usage = self._format_usage(command) sections.append(f"Usage:\n{self.indent}{usage}") # Arguments if args := command.get('arguments'): section = self._format_section('Arguments', args) sections.append(section) # Options if opts := command.get('options'): section = self._format_section('Options', opts) sections.append(section) # Examples if examples := command.get('examples'): section = self._format_examples(examples) sections.append(section) return '\n\n'.join(sections) def _format_section(self, title: str, items: List[dict]) -> str: lines = [f"{title}:"] for item in items: name = item.get('name') or item.get('flags', '') desc = item.get('description', '') lines.append(f"{self.indent}{name:<20} {desc}") return '\n'.join(lines) def _format_examples(self, examples: List[dict]) -> str: lines = ["Examples:"] for ex in examples: lines.append(f"{self.indent}$ {ex['command']}") lines.append(f"{self.indent} {ex['description']}") lines.append('') return '\n'.join(lines).rstrip() def _format_usage(self, command: dict) -> str: parts = [command['name']] for arg in command.get('arguments', []): if arg.get('required', True): parts.append(f"<{arg['name']}>") else: parts.append(f"[{arg['name']}]") parts.append('[options]') return ' '.join(parts)
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 19,095 | 15,419 | -19% | 1 | 1 | 0% | 4,537 | 4,988 | +10% | 0 | 0 | — |
case-02 | fail→pass | 19,358 | 14,504 | -25% | 1 | 1 | 0% | 4,594 | 4,680 | +2% | 0 | 0 | — |
case-03 | pass→pass | 15,732 | 13,713 | -13% | 1 | 1 | 0% | 3,582 | 4,624 | +29% | 0 | 0 | — |
case-04 | fail→fail | 14,268 | 19,187 | +34% | 1 | 1 | 0% | 2,834 | 4,752 | +68% | 0 | 0 | — |
case-05 | pass→pass | 8,909 | 8,134 | -9% | 1 | 1 | 0% | 1,921 | 3,114 | +62% | 0 | 0 | — |
case-06 | fail→pass | 12,190 | 5,614 | -54% | 1 | 1 | 0% | 2,205 | 2,511 | +14% | 0 | 0 | — |
case-07 | pass→pass | 16,650 | 11,811 | -29% | 1 | 1 | 0% | 2,779 | 4,132 | +49% | 0 | 0 | — |
case-08 | pass→pass | 8,878 | 5,763 | -35% | 1 | 1 | 0% | 1,892 | 2,594 | +37% | 0 | 0 | — |
case-09 | fail→fail | 10,936 | 10,473 | -4% | 1 | 1 | 0% | 2,248 | 3,837 | +71% | 0 | 0 | — |
case-10 | fail→fail | 12,339 | 7,402 | -40% | 1 | 1 | 0% | 2,215 | 3,040 | +37% | 0 | 0 | — |
case-11 | fail→pass | 17,857 | 15,020 | -16% | 1 | 1 | 0% | 3,904 | 4,749 | +22% | 0 | 0 | — |
case-12 | fail→fail | 11,575 | 11,869 | +3% | 1 | 1 | 0% | 2,096 | 3,887 | +85% | 0 | 0 | — |
case-13 | fail→fail | 8,780 | 4,243 | -52% | 1 | 1 | 0% | 1,584 | 2,277 | +44% | 0 | 0 | — |
case-14 | fail→pass | 16,953 | 17,807 | +5% | 1 | 1 | 0% | 3,012 | 4,648 | +54% | 0 | 0 | — |
case-15 | fail→fail | 9,832 | 7,836 | -20% | 1 | 1 | 0% | 2,012 | 3,067 | +52% | 0 | 0 | — |
case-16 | fail→fail | 14,423 | 13,922 | -3% | 1 | 1 | 0% | 3,198 | 4,459 | +39% | 0 | 0 | — |
case-17 | fail→pass | 5,834 | 4,522 | -22% | 1 | 1 | 0% | 1,173 | 2,116 | +80% | 0 | 0 | — |
case-18 | fail→fail | 15,140 | 12,181 | -20% | 1 | 1 | 0% | 2,636 | 3,540 | +34% | 0 | 0 | — |
case-19 | fail→fail | 12,017 | 10,615 | -12% | 1 | 1 | 0% | 1,981 | 3,071 | +55% | 0 | 0 | — |
case-20 | fail→fail | 6,942 | 7,929 | +14% | 1 | 1 | 0% | 1,683 | 3,261 | +94% | 0 | 0 | — |
case-21 | fail→fail | 15,052 | 9,513 | -37% | 1 | 1 | 0% | 2,883 | 3,543 | +23% | 0 | 0 | — |
case-22 | fail→fail | 8,139 | 10,808 | +33% | 1 | 1 | 0% | 2,012 | 3,989 | +98% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +23 percentage points is the difference between those two pass rates over the 22 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.