TODO: [What it does]. Use when [specific triggers]. Example: "Analyzes Excel spreadsheets and generates charts. Use when working with Excel files, .xlsx, spreadsheet analysis, or data visualization." REQUIRED: Include "Use when..." with trigger keywords (file types, domains, tasks). Third person. Max 1024 chars.
Inherits all available tools
Additional assets for this skill
This skill inherits all available tools. When active, it can use any tool Claude has access to.
assets/example.txtreferences/example.mdscripts/example.pyNOTE: This is a template file. After copying this directory to create a new skill, customize the YAML frontmatter and replace all TODO sections with your skill's content.
TODO: Brief 1-2 sentence overview. Be concise - the context window is shared with conversation history, other skills, and user requests.
TODO: Detailed introduction explaining the purpose and scope of this skill (1 paragraph).
This skill should be used when:
TODO: Provide the fastest path to value. Show the most common use case with a brief example.
Example:
# TODO: Replace with actual quick start command or code snippet
echo "Hello, world!"
TODO: Choose and implement ONE of these structural patterns based on your skill's purpose:
Use when: Multi-step processes, sequential operations, guided workflows
Pattern structure:
## Workflow Decision Tree
## Step 1: Initial Setup
## Step 2: Configuration
## Step 3: Execution
## Troubleshooting
For complex workflows, provide a checklist:
## PDF Form Filling Workflow
Copy this checklist and check off items as you complete them:
```
Task Progress:
- [ ] Step 1: Analyze the form
- [ ] Step 2: Create field mapping
- [ ] Step 3: Validate mapping
- [ ] Step 4: Fill the form
- [ ] Step 5: Verify output
```
**Step 1: Analyze the form**
[Instructions for this step]
**Step 2: Create field mapping**
[Instructions for this step]
[Continue for each step...]
Use when: Tool collections, utility skills, multiple independent capabilities
Pattern structure:
## Task Category 1: [Name]
### Task 1.1: [Operation]
### Task 1.2: [Operation]
## Task Category 2: [Name]
### Task 2.1: [Operation]
Use when: Style guides, coding standards, brand guidelines, API specifications
Pattern structure:
## Core Principles
## Guidelines
## Specifications
## Usage Examples
Use when: Complex systems, integrated tools, multi-capability skills
Pattern structure:
## Core Capabilities
## Capability 1: [Feature Name]
## Capability 2: [Feature Name]
## Integration Guide
Use when: Complex operations, batch updates, operations where errors are costly
Pattern structure:
## Workflow with Validation
### Step 1: Analyze Input
[Understand requirements]
### Step 2: Generate Plan
Create intermediate plan file (e.g., changes.json)
### Step 3: Validate Plan
```bash
python scripts/validate_plan.py plan.json
```
### Step 4: Review Errors
If errors found, fix and return to Step 3
### Step 5: Execute Plan
Apply validated plan
### Step 6: Verify Output
Confirm output meets requirements
Key principle: Catch errors early through intermediate validation before expensive operations.
TODO: Provide concrete, representative examples with input/output pairs. Examples help Claude understand desired style and level of detail better than descriptions alone.
TODO: Show a simple, common use case. Use input/output format:
Input: [What the user provides]
Expected Output:
[Show exactly what should be produced]
Format guidance: Show concrete input → output pairs. For instance, if this skill generates commit messages, show:
TODO: Show a more complex or powerful use case using the same input/output format:
Input: [More complex scenario]
Expected Output:
[Show the complete output with all details]
Pattern tip: Use input/output pairs like in regular prompting to show Claude the desired format and quality level. This works better than descriptions alone.
This skill includes example resource directories that demonstrate how to organize different types of bundled resources:
Executable code (Python/Bash/etc.) that can be run directly to perform specific operations.
Examples from other skills:
fill_fillable_fields.py, extract_form_field_info.py - utilities for PDF manipulationdocument.py, utilities.py - Python modules for document processingAppropriate for: Python scripts, shell scripts, or any executable code that performs automation, data processing, or specific operations.
Note: Scripts may be executed without loading into context, but can still be read by Claude for patching or environment adjustments.
TODO: If your skill includes scripts, list them here:
scripts/example.py - TODO: Describe what this script doesDocumentation and reference material intended to be loaded into context to inform Claude's process and thinking.
Examples from other skills:
communication.md, context_building.md - detailed workflow guidesAppropriate for: In-depth documentation, API references, database schemas, comprehensive guides, or any detailed information that Claude should reference while working.
IMPORTANT: Keep references one level deep from SKILL.md - don't nest references beyond one level. Claude reads complete files when directly referenced from SKILL.md, but may only preview nested references.
For large reference files (>100 lines): Include a table of contents at the top so Claude understands available information during partial reads.
TODO: If your skill includes reference documentation, list it here:
Make content greppable:
To find OAuth implementation details:
```bash
grep -i "oauth" references/authentication.md
```
Files not intended to be loaded into context, but rather used within the output Claude produces.
Examples from other skills:
Appropriate for: Templates, boilerplate code, document templates, images, icons, fonts, or any files meant to be copied or used in the final output.
TODO: If your skill includes assets, list them here:
assets/example.txt - TODO: Describe what this asset is forAny unneeded directories can be deleted. Not every skill requires all three types of resources.
TODO: Document common issues and solutions:
Issue: [Common Problem]
Issue: [Another Common Problem]
TODO: List recommendations for using this skill effectively:
Before deploying this skill, complete all TODO items:
skill-name in frontmatter with actual name (lowercase, hyphens, max 64 chars)description with both functionality AND trigger scenarios (max 1024 chars)name field exactlyStructural Pattern Selection:
Content Sizing:
The context window is shared with conversation history, other skills, and user requests.
Set Appropriate Degrees of Freedom:
Match specificity to task fragility:
Description Tips:
CRITICAL: Always use third person - Description is injected into system prompt:
Include these keyword types:
.md, .json, .xlsx, PDF, ExcelAPI, authentication, database, testinganalyze, generate, create, build, validateGit, Docker, Kubernetes, PostgreSQLOfficial Pattern: [What it does]. Use when [specific triggers].
This is the ONLY documented mechanism for skill discovery. Claude uses descriptions to choose skills from 100+ available. Include BOTH capabilities AND triggers.
Example descriptions:
✅ Excellent (follows official pattern):
description: Analyzes Excel spreadsheets, generates pivot tables, creates charts. Use when working with Excel files (.xlsx, .xls), spreadsheet analysis, or data visualization tasks.
✅ Good (clear triggers):
description: Generates descriptive commit messages by analyzing git diffs. Use when writing commit messages, reviewing staged changes, or preparing git commits.
❌ Missing "Use when..." triggers:
description: Generates descriptive commit messages.
❌ Too vague:
description: Helps with files.
❌ Wrong voice (must be third person):
description: I will help you analyze spreadsheets.
Validation Checklist:
name follows conventions (lowercase, hyphens, max 64 chars)name matches directory name exactlydescription follows official pattern: [What it does]. Use when [triggers].description includes "Use when..." with specific trigger keywords