Skip to content
Back to skills

2651 023 Ms Oldv Plugin Repo 0b2ef9f4

ASecurity

Generated: 2025-10-11 Repository: `jeremylongshore/claude-code-plugins` Location: `/home/jeremy/projects/claude-code-plugins` ---

  • 9 stars
  • 0 votes
  • 0 copies
  • 1 view
  • Added October 11, 2026
devopsjavascripttypescriptjavashellbashreactvuenodedockerkubernetes

Works with

  • claude code
  • api
  • mcp

Security analysis

A100/100

Scanned October 11, 2026

npx -y skills add tools-only/X-Skills --skill 2651-023-ms-oldv-plugin-repo_0b2ef9f4 --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of 2651 023 Ms Oldv Plugin Repo 0b2ef9f4?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for 2651 023 Ms Oldv Plugin Repo 0b2ef9f4
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/tools-only-2651-023-ms-oldv-plugin-repo-0b2ef9f4/badge)](https://www.skillsdirectory.com/skills/tools-only-2651-023-ms-oldv-plugin-repo-0b2ef9f4)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

SKILL.md
# Plugin Repository Report

Generated: 2025-10-11
Repository: `jeremylongshore/claude-code-plugins`
Location: `/home/jeremy/projects/claude-code-plugins`

---

## 1. Repository Structure

### Compact Tree (Depth 4)
```
claude-code-plugins/
├── .claude-plugin/
│   └── marketplace.json (10118 bytes)
├── .github/
│   ├── ISSUE_TEMPLATE/
│   ├── workflows/
│   │   ├── deploy-marketplace.yml
│   │   ├── release.yml
│   │   └── validate-plugins.yml
│   ├── FUNDING.yml
│   ├── PULL_REQUEST_TEMPLATE.md
│   └── RELEASE_CHECKLIST.md
├── plugins/
│   ├── ai-agency/          # 6 AI agency plugins
│   ├── community/          # (empty - for community contributions)
│   ├── devops/             # Production DevOps plugins
│   ├── examples/           # 3 example plugins
│   ├── mcp/                # 5 MCP server plugins
│   ├── packages/           # 4 plugin packs
│   └── productivity/       # Productivity plugins
├── marketplace/            # Astro-based marketplace website
├── docs/                   # Documentation
├── scripts/                # Repository maintenance scripts
├── templates/              # 4 plugin templates
└── claudes-docs/          # Generated documentation
```

### Plugin Packages Structure
```
packages/
├── ai-ml-engineering-pack/    # 12 AI/ML plugins
├── devops-automation-pack/    # 25 DevOps plugins
├── fullstack-starter-pack/    # 15 fullstack plugins
└── security-pro-pack/          # 10 security plugins
```

### MCP Plugins
```
mcp/
├── conversational-api-debugger/
├── design-to-code/
├── domain-memory-agent/
├── project-health-auditor/
└── workflow-orchestrator/
```

---

## 2. Plugin Metadata

### Main Marketplace Configuration
| Field | Value |
|-------|-------|
| Name | claude-code-plugins |
| Owner | Jeremy Longshore |
| Version | 1.1.0 |
| Homepage | https://claudecodeplugins.io/ |
| Plugins Count | 20 registered in marketplace.json |

### Sample Plugin Pack Metadata (DevOps Automation)
| Field | Value |
|-------|-------|
| Name | devops-automation-pack |
| Version | 1.0.0 |
| Description | 25 professional DevOps plugins for CI/CD, deployment, Docker, Kubernetes... |
| Author | Jeremy Longshore <jeremy@intentsolutions.ai> |
| License | MIT |
| Keywords | devops, ci-cd, docker, kubernetes, terraform, deployment |

---

## 3. MCP Integrations

### Discovered MCP Servers (5 total)

| Plugin | MCP Config | Server Name | Type |
|--------|------------|-------------|------|
| project-health-auditor | .mcp.json | code-metrics | node |
| conversational-api-debugger | .mcp.json | (present) | node |
| domain-memory-agent | .mcp.json | (present) | node |
| design-to-code | .mcp.json | (present) | node |
| workflow-orchestrator | .mcp.json | (present) | node |

### MCP Server Configuration Example
```json
{
  "mcpServers": {
    "code-metrics": {
      "command": "node",
      "args": ["${CLAUDE_PLUGIN_ROOT}/dist/servers/code-metrics.js"],
      "env": {}
    }
  }
}
```

**Note**: MCP tools are defined in compiled JavaScript files in `dist/servers/` directories.

---

## 4. Slash Commands

### Example Commands Found

| Plugin | Command | Description |
|--------|---------|-------------|
| project-health-auditor | /analyze | Trigger full repository health analysis |
| git-commit-smart | (in commands/) | AI-powered commit messages |
| n8n-workflow-designer | (in commands/) | Design n8n workflows |
| make-scenario-builder | (in commands/) | Build Make.com scenarios |
| zapier-zap-builder | (in commands/) | Create Zapier Zaps |

### Command Structure Example (project-health-auditor/analyze.md)
- **Name**: Repository Health Analysis
- **MCP Calls**: `list_repo_files`, `file_metrics`, `git_churn`, `map_tests`
- **Workflow**:
  1. File Discovery
  2. Complexity Analysis
  3. Git Churn Analysis
  4. Test Coverage Mapping
- **Output**: Health report with actionable recommendations

---

## 5. Sub-agents

### Discovered Agents (Sample)

| Pack | Agent | Purpose |
|------|-------|---------|
| ai-ml-engineering | prompt-optimizer | Optimize prompts for efficiency |
| ai-ml-engineering | llm-integration-expert | Multi-provider LLM integration |
| ai-ml-engineering | rag-architect | Design RAG systems |
| security-pro | compliance-checker | Check HIPAA, GDPR, PCI DSS compliance |
| security-pro | crypto-expert | Cryptography auditing |
| devops-automation | ci-cd-expert | CI/CD pipeline optimization |
| fullstack-starter | frontend-architect | React/Vue component design |

### Agent Count by Pack
- **AI/ML Engineering**: 8 agents
- **Security Pro**: 5 agents
- **DevOps Automation**: 5 agents
- **Fullstack Starter**: 6 agents

---

## 6. Hooks

### Discovered Hook Implementations

| Plugin | Trigger | Action | File Modified |
|--------|---------|--------|---------------|
| formatter | PostToolUse (Write\|Edit) | format.sh | Formats written/edited files |

### Hook Structure Example
```json
{
  "hooks": {
    "PostToolUse": [{
      "matcher": "Write|Edit",
      "hooks": [{
        "type": "command",
        "command": "${CLAUDE_PLUGIN_ROOT}/scripts/format.sh"
      }]
    }]
  }
}
```

---

## 7. GitHub Automation

### Workflow Files

#### validate-plugins.yml
- **Triggers**: Pull requests (plugins/**, .claude-plugin/**), Push to main
- **Main Steps**:
  - Checkout code
  - Validate JSON files with jq
  - Check plugin structure (plugin.json, README.md)
  - Verify scripts are executable
- **Env/Secrets**: None required

#### release.yml
- **Triggers**: Manual dispatch, tags
- **Main Steps**:
  - Build release artifacts
  - Create GitHub release
  - Publish to marketplace
- **Env/Secrets**: GITHUB_TOKEN

#### deploy-marketplace.yml
- **Triggers**: Push to main (marketplace changes)
- **Main Steps**:
  - Build Astro site
  - Deploy to GitHub Pages
- **Env/Secrets**: GITHUB_TOKEN

---

## 8. Quick Lint

### JSON Syntax Check
-  marketplace.json: Valid
-  plugin.json files: Valid (sampled)
-  hooks.json files: Valid
-  .mcp.json files: Valid

### Command File Validation
- ️ Command files should have H1 matching slash command name
-  Commands reference existing MCP tools (where applicable)

### Script Permissions
-  Shell scripts in examples/formatter/scripts/ are present
- ️ Need to verify all .sh files have execute permissions

---

## 9. Usage Snippet

### Installation

```bash
# Add the marketplace to Claude Code
/plugin marketplace add jeremylongshore/claude-code-plugins

# Install a plugin pack
/plugin install devops-automation-pack@claude-code-plugins-plus

# Install individual plugins
/plugin install project-health-auditor@claude-code-plugins-plus
/plugin install git-commit-smart@claude-code-plugins-plus
```

### Example Invocations

```bash
# Analyze repository health
/analyze /path/to/repo

# Generate smart commit message
/gc

# Design n8n workflow
/n8n-workflow

# Trigger security audit
/security-audit
```

### Local Testing

```bash
# Create test marketplace
mkdir -p ~/test-marketplace/.claude-plugin
cat > ~/test-marketplace/.claude-plugin/marketplace.json << 'EOF'
{
  "name": "test",
  "owner": {"name": "Test"},
  "plugins": [{
    "name": "my-plugin",
    "source": "/absolute/path/to/plugin"
  }]
}
EOF

# Add and install
/plugin marketplace add ~/test-marketplace
/plugin install my-plugin@test
```

---

## 10. Gaps & TODOs

### Concrete Fixes Needed

1. **Missing plugin.json files**
   - [ ] Add `.claude-plugin/plugin.json` to individual example plugins
   - [ ] Add `.claude-plugin/plugin.json` to AI agency plugins

2. **MCP Server Configuration**
   - [ ] Document MCP tools in README for each MCP plugin
   - [ ] Add tool descriptions to plugin documentation

3. **Command Documentation**
   - [ ] Ensure all command files have proper frontmatter
   - [ ] Verify command names match file names

4. **Script Permissions**
   - [ ] Run `chmod +x` on all .sh files in scripts/ directories
   - [ ] Add permission check to validation script

5. **Agent Documentation**
   - [ ] Add frontmatter to all agent .md files
   - [ ] Document expected inputs/outputs

6. **Hook Implementations**
   - [ ] Add hooks examples to more plugins
   - [ ] Document hook side effects clearly

7. **Package Structure**
   - [ ] Standardize sub-plugin structure in packages
   - [ ] Add individual plugin.json to each sub-plugin

8. **Testing**
   - [ ] Add test scripts for MCP servers
   - [ ] Create integration tests for commands

---

## Summary

**Repository Stats:**
- Total Plugins: 20 in marketplace
- Plugin Packs: 4 (62 total components)
- MCP Servers: 5 (with compiled TypeScript)
- Example Plugins: 3
- Templates: 4
- GitHub Workflows: 3

**Key Features:**
- Comprehensive marketplace with installation system
- Professional plugin packs for different domains
- MCP server integration for external tools
- Automated validation and release workflows
- Active development with clear structure

**Repository Health:**  Well-organized, actively maintained, ready for community contributions

---

*Report generated using Claude Code analysis tools*

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.