Claude Code Plugins

Community-maintained marketplace

Feedback

contribute-skill

@transilienceai/communitytools
4
0

End-to-end workflow automation for contributing new skills to the repository

Install Skill

1Download skill
2Enable skills in Claude

Open claude.ai/settings/capabilities and find the "Skills" section

3Upload to Claude

Click "Upload skill" and select the downloaded ZIP file

Note: Please verify skill by going through its instructions before using it.

SKILL.md

name contribute-skill
description End-to-end workflow automation for contributing new skills to the repository
tools Read, Write, Bash
model inherit
max_turns 10
max_budget 0.2

Contribute Skill Agent

Purpose

End-to-end workflow automation for contributing a completely new skill to the security repository. This agent orchestrates the entire process from GitHub issue creation to PR submission.

When to Use

  • User wants to contribute a brand new skill
  • User says: "I want to add a skill for [purpose]"
  • User wants to create a new security analysis capability

Workflow Steps

Step 1: Gather Skill Information

Ask the user for the following information:

  1. Skill Name (e.g., "AWS CloudTrail Analyzer")
    • Will be converted to directory name: aws_cloudtrail_analyzer
  2. Skill Purpose (1-2 sentences describing what it does)
  3. Category (choose one):
    • Compliance (PCI-DSS, SOC2, NIST, etc.)
    • Penetration Testing
    • Incident Response
    • Vulnerability Assessment
    • Cloud Security (AWS/Azure/GCP)
    • Other (specify)
  4. Cloud Provider (if applicable): AWS, Azure, GCP, Multi-cloud, None
  5. Key Features (list 3-5 main capabilities)
  6. Example Agents (what agents should be included? e.g., log-parser, finding-analyzer)

Step 2: Create GitHub Issue

Use the git-issue-creator agent to create a skill proposal issue:

  • Title: "New Skill: [Skill Name]"
  • Body: Use skill proposal template
  • Labels: enhancement, new-skill, [category]
  • Capture the issue number (e.g., #123)

Command:

gh issue create \
  --title "New Skill: [Skill Name]" \
  --body "$(cat templates/issue_templates/skill_proposal.md)" \
  --label "enhancement,new-skill,[category]"

Step 3: Create Git Branch

Use the git-branch-manager agent:

  • Ensure we're on the main branch and up to date
  • Create feature branch: feature/[skill-directory-name]
  • Switch to the new branch

Commands:

git checkout main
git pull origin main
git checkout -b feature/[skill-directory-name]

Step 4: Generate Skill Structure

Use the skill-generator agent:

  • Create skill directory: custom_skills/[skill-directory-name]/
  • Generate all required files from templates
  • Populate with user-provided information

Files to create:

  1. SKILL.md - From skill template, populated with skill details
  2. CLAUDE.md - Basic context file
  3. README.md - User-facing documentation
  4. .claude/agents/ directory
  5. Example agent files (e.g., example-analyzer.md)
  6. tools/ directory
  7. scripts/ directory
  8. outputs/.gitkeep file
  9. templates/ directory (optional)
  10. reference/ directory (optional)

Step 5: Populate Files with Content

Use the documentation-writer agent to generate content:

SKILL.md content:

---
name: [skill-directory-name]
description: [Skill purpose from user input]
---

# [Skill Name]

## Purpose
[Detailed purpose and use cases]

## Category
[Category from user input]

## Key Features
[List of features from user input]

## Agents
[List of example agents]

## Usage
[Basic usage instructions]

README.md content:

# [Skill Name]

[Skill purpose]

## Features
- [Feature 1]
- [Feature 2]
- [Feature 3]

## Getting Started
[Basic setup and usage]

## Examples
[Usage examples]

## Contributing
See the main repository [CONTRIBUTING.md](../../CONTRIBUTING.md)

Example Agent (.claude/agents/example-agent.md):

# [Agent Name]

## Purpose
[What this agent does]

## Inputs
[What information/data this agent needs]

## Outputs
[What this agent produces]

## Process
[Step-by-step what this agent does]

Step 6: Commit Changes

Use conventional commit format:

git add custom_skills/[skill-directory-name]/
git commit -m "feat([skill-directory-name]): Add [Skill Name] skill

- Create initial skill structure
- Add SKILL.md and README.md
- Create example agents
- Set up directory structure

Fixes #[issue-number]"

Step 7: Push Branch

Push the feature branch to remote:

git push -u origin feature/[skill-directory-name]

Step 8: Create Pull Request

Use the git-pr-creator agent:

  • Title: "feat([skill-directory-name]): Add [Skill Name] skill"
  • Body: Use PR template with checklist
  • Link to issue: "Closes #[issue-number]"
  • Add labels: enhancement, new-skill
  • Request review if configured

Command:

gh pr create \
  --title "feat([skill-directory-name]): Add [Skill Name] skill" \
  --body "$(cat templates/pr_template.md)" \
  --label "enhancement,new-skill" \
  --base main

Step 9: Provide Summary

Display completion summary:

✓ Skill contribution complete!

Summary:
- Issue: #[issue-number] - [Issue URL]
- Branch: feature/[skill-directory-name]
- PR: #[pr-number] - [PR URL]
- Files created:
  ✓ SKILL.md
  ✓ README.md
  ✓ CLAUDE.md
  ✓ [X] agent files
  ✓ Directory structure

Next Steps:
1. Review the PR at: [PR URL]
2. Make any additional changes by pushing to the same branch
3. Wait for review feedback
4. Address any review comments

To make changes:
  git checkout feature/[skill-directory-name]
  [make your changes]
  git add .
  git commit -m "refactor: Update based on feedback"
  git push

Error Handling

If issue creation fails:

  • Check GitHub CLI is installed: gh --version
  • Check authentication: gh auth status
  • Provide fallback instructions for manual issue creation

If branch already exists:

  • Ask user if they want to:
    1. Delete and recreate
    2. Continue with existing branch
    3. Choose a different branch name

If files already exist:

  • Warn user that skill directory exists
  • Ask if they want to:
    1. Overwrite
    2. Cancel
    3. Choose different skill name

If git operations fail:

  • Check git status
  • Ensure no uncommitted changes on main
  • Check remote connectivity
  • Provide clear error messages with resolution steps

Dependencies

This agent requires:

  • Git installed and configured
  • GitHub CLI (gh) installed and authenticated
  • Write access to the repository
  • Network connectivity

Check dependencies:

git --version
gh --version
gh auth status

Agent Invocations

This agent should invoke these supporting agents:

  1. git-issue-creator - For GitHub issue creation
  2. git-branch-manager - For branch operations
  3. skill-generator - For file/directory creation
  4. documentation-writer - For content generation
  5. git-pr-creator - For PR creation

Example Usage

User Input: "I want to contribute a skill for analyzing AWS VPC Flow Logs for security incidents"

Agent Actions:

  1. Gathers details:

    • Name: "AWS VPC Flow Log Analyzer"
    • Directory: aws_vpc_flow_analyzer
    • Category: Incident Response
    • Cloud: AWS
    • Features: Parse flow logs, detect anomalies, identify threats, timeline analysis
    • Agents: log-parser, anomaly-detector, threat-analyzer

  2. Creates issue #145

  3. Creates branch: feature/aws-vpc-flow-analyzer

  4. Generates complete skill structure

  5. Commits and pushes

  6. Creates PR #146

  7. Provides summary with links

Output Validation

Before completing, verify:

  • GitHub issue created successfully
  • Branch created and checked out
  • All required files exist
  • SKILL.md has proper YAML frontmatter
  • README.md is complete
  • At least one example agent exists
  • Commit follows conventional format
  • Branch pushed to remote
  • PR created and linked to issue
  • User provided with all necessary links

Notes

  • Always use conventional commit format
  • Always link PRs to issues with "Fixes #" or "Closes #"
  • Use descriptive commit messages
  • Follow repository naming conventions
  • Ensure all templates are properly populated
  • Provide clear next steps to the user