/basic-command-template

Basic Command Template

Use this template as a starting point for creating new Claude Code commands.

File: `.claude/commands/command-name.md`

---
description: Brief description of what this command does (keep under 80 characters)
category: category-name
argument-hint: <expected-arguments>
allowed-tools: Read, Write, Bash, Grep, Glob
model: claude-sonnet-4-6
---

# [Command Name]

You are a [Role/Expert Type] specializing in [Domain/Skill Area].

## Your Task

When the user invokes this command with `/command-name [arguments]`:

1. **Analyze** the provided [input type]
2. **Process** according to [specific criteria]
3. **Generate** [expected output format]
4. **Validate** results for [quality measures]

## Input Requirements

- **Required**: [What must be provided]
- **Optional**: [What can be provided]
- **Format**: [Expected input format]

## Processing Steps

### Step 1: Validation
- Check that [validation criteria]
- Verify [requirements]
- Handle edge cases like [specific examples]

### Step 2: Analysis
- [Specific analysis approach]
- Focus on [key aspects]
- Consider [important factors]

### Step 3: Output Generation
- Create [output type] that includes:
  - [Required element 1]
  - [Required element 2]
  - [Required element 3]

## Output Format

[Specific structure example]

## Examples

### Example 1: [Scenario Name]

/command-name input-example

**Expected Output:**

[Sample output showing expected format]

### Example 2: [Edge Case]

/command-name --option value

**Expected Output:**

[Sample output for edge case]

## Error Handling

If the command encounters issues:
- **Invalid input**: Explain what's wrong and suggest corrections
- **Missing files**: List what files are needed and where to find them
- **Permission errors**: Indicate what permissions are required
- **Unexpected errors**: Provide debugging information and next steps

## Success Criteria

This command succeeds when:
- [ ] All required inputs are validated
- [ ] Processing completes without errors
- [ ] Output follows specified format
- [ ] Results meet quality standards
- [ ] User receives actionable information

Customization Guide

1. Replace Placeholders

| Placeholder | Replace With | Example | |-------------|--------------|---------| | `[Command Name]` | Actual command name | "Code Reviewer" | | `[Role/Expert Type]` | Expertise area | "Senior Code Reviewer" | | `[Domain/Skill Area]` | Specialization | "security and performance" | | `[input type]` | Expected input | "source code files" | | `[specific criteria]` | Processing rules | "OWASP security guidelines" | | `[expected output format]` | Result format | "structured review report" | | `[quality measures]` | Success metrics | "actionable feedback" |

2. Configure Frontmatter

# Model selection based on complexity — pin specific variants, not aliases.
# Bare aliases (`sonnet`, `opus`, `haiku`) inherit the parent session's
# variant, which under a 1M-context parent triggers the usage-credit gate
# and blocks subagent dispatch. See aiwg #1442.
model: claude-haiku-4-5      # For simple, fast operations
model: claude-sonnet-4-6     # For balanced tasks (recommended default)
model: claude-opus-4-7       # For complex reasoning tasks

# Tool selection based on needs
allowed-tools: Read                    # Read-only commands
allowed-tools: Read, Write             # File manipulation
allowed-tools: Bash, Read, Write       # System interaction
allowed-tools: Read, Grep, Glob        # Search and analysis

# Category organization (choose one)
category: sdlc-management              # SDLC workflow commands
category: security-quality             # Security and quality gates
category: documentation-tracking       # Documentation and traceability
category: project-task-management      # Project planning and management
category: version-control-git          # Git and version control
category: code-analysis-testing        # Testing and code analysis
category: team-collaboration           # Team and retrospective tools
category: documentation-changelogs     # Documentation generation

3. Add Domain-Specific Logic

## Domain-Specific Considerations

### For Code Analysis Commands
- Include language-specific patterns
- Reference relevant style guides
- Consider framework conventions
- Add security vulnerability checks

### For Documentation Commands
- Follow documentation standards
- Include examples and usage
- Consider audience expertise level
- Ensure searchable structure

### For Infrastructure Commands
- Include safety checks
- Validate configurations
- Consider environment differences
- Add rollback procedures

Validation Checklist

Before deploying your command:

• [ ] **Functionality**: Command performs intended task correctly
• [ ] **Error Handling**: Gracefully handles invalid inputs and edge cases
• [ ] **Documentation**: Clear instructions and examples provided
• [ ] **Security**: No unnecessary tool permissions granted
• [ ] **Performance**: Uses appropriate model for task complexity
• [ ] **Testing**: Verified with multiple test cases
• [ ] **Integration**: Works well with existing commands and agents

Testing Your Command

1. Basic Functionality Test

# Test with typical input
/your-command normal-input

# Test with edge cases
/your-command ""
/your-command --help

2. Error Condition Tests

# Test with missing arguments
/your-command

# Test with invalid input
/your-command invalid-input

3. Integration Tests

# Test with other commands
/your-command | /another-command

# Test in workflow context

Common Patterns

File Processing Command

## Your Task
1. **Read** the specified file using the Read tool
2. **Analyze** content for [specific criteria]
3. **Transform** according to [rules]
4. **Write** results to [output location]

Analysis Command