opencode-skills-maintainer-skill

What I do

I maintain skill consistency, quality, and efficiency by:

1. Discover All Skills: Scan the skills/ folder to discover all available skills
2. Extract Skill Metadata: Read frontmatter from each SKILL.md file (name, description, category)
3. Validate Skill Structure: Ensure all skills have required fields and valid frontmatter
4. Categorize Skills: Organize skills into logical categories (Framework, Test Generators, Linters, etc.)
5. Detect Redundancy: Identify overlapping functionality, duplicate capabilities, and consolidation opportunities
6. Analyze Modularization: Recommend skill decomposition and reusable component extraction
7. Generate Report: Provide comprehensive summary with validation status and optimization recommendations

When to use me

Use this skill when:

• You want to audit all skills in the repository
• You need to validate skill metadata consistency
• You're checking for missing required fields in SKILL.md files
• You want a categorized list of all available skills
• You're debugging skill discovery issues
• You need to identify redundant functionality across multiple skills
• You're planning to refactor or consolidate the skill library
• You want to improve maintainability and reduce code duplication in skills

Prerequisites

• Access to the repository root directory
• jq tool installed for JSON validation
• Python 3+ installed for YAML parsing

Steps

Step 1: Discover All Skills

Scan the skills/ folder to find all skill directories:

# Find all skill directories
find skills/ -name "SKILL.md" -type f | sort

Step 2: Extract Skill Metadata

For each skill, read the frontmatter to extract:

# Extract name and description from frontmatter
for skill_dir in skills/*/; do
  echo "=== $(basename "$skill_dir") ==="
  head -10 "$skill_dir/SKILL.md" | grep -E "(name:|description:)" | head -2
  echo
done

Required Fields:

• name: The skill identifier
• description: Brief description of what the skill does
• Optional: category, workflow, audience (from metadata section)