Ghost Security MCP Server
OfficialClick on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Ghost Security MCP Server show my security findings for the main repository"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Ghost Security MCP Server
A Model Context Protocol (MCP) server for the Ghost Security API, providing secure access to security findings and repository data through standardized tools.
Features
Security Findings Management: Query, retrieve, and update security findings
Repository Analysis: Access repository data and associated findings
Universal MCP Compatibility: Works with any MCP-compatible client or application
Bundled CLI Chatbot: Includes a simple chatbot for quick interaction
Comprehensive API Coverage: Full support for Ghost Security API (V2) endpoints
Type Safety: Complete TypeScript implementation with proper typing
Error Handling: Robust error handling and validation
Related MCP server: CyberMCP
Installation
Quick Install for Claude Code CLI
Option 1: Interactive Installer (Easiest)
# Clone and build
git clone https://github.com/ghostsecurity/ghost-mcp-server
cd ghost-mcp-server
npm install && npm run build
# Run interactive installer (prompts for API key)
npm run install:interactiveOption 2: Automated Setup Script
# Clone and build
git clone https://github.com/ghostsecurity/ghost-mcp-server
cd ghost-mcp-server
npm install && npm run build
# Run setup script with your API key
./setup-claude-code.sh YOUR_API_KEY
# Optional: Scope to a specific repository
./setup-claude-code.sh YOUR_API_KEY YOUR_REPO_IDOption 3: Manual Setup
# Clone and build
git clone https://github.com/ghostsecurity/ghost-mcp-server
cd ghost-mcp-server
npm install && npm run build
# Add to Claude Code
claude mcp add --transport stdio --scope user ghost-security -- node $(pwd)/dist/index.js
# Then edit ~/.claude.json and add under mcpServers.ghost-security.env:
# {
# "GHOST_SECURITY_API_KEY": "your-api-key",
# "GHOST_SECURITY_BASE_URL": "https://api.ghostsecurity.ai/v2"
# }Option 4: From npm (Coming Soon)
Once published to npm, installation will be even simpler:
claude mcp add --transport stdio --scope user ghost-security -- npx ghostsecurity-mcpThen add your API key to ~/.claude.json:
{
"mcpServers": {
"ghost-security": {
"env": {
"GHOST_SECURITY_API_KEY": "your-api-key-here",
"GHOST_SECURITY_BASE_URL": "https://api.ghostsecurity.ai/v2"
}
}
}
}Standard Installation (All MCP Clients)
Clone this repository:
git clone <repository-url>
cd ghost-mcp-serverInstall dependencies:
npm installBuild the project:
npm run buildUsage
Configuration
The server requires a Ghost Security API key to be provided at startup. You can provide it in two ways:
Option 1: Environment Variable (Recommended)
export GHOST_SECURITY_API_KEY="your-ghost-security-api-key"
export GHOST_SECURITY_BASE_URL="https://api.ghostsecurity.ai/v2" # optional, defaults to v2
export GHOST_SECURITY_REPO_ID="your-repository-id" # optional - scope to specific repoOption 2: Command Line Argument
node dist/index.js "your-ghost-security-api-key" "your-repository-id"Note: Repository ID is optional. If provided, all operations will be scoped to that repository.
Running the Server
For development:
# Set environment variable first
export GHOST_SECURITY_API_KEY="your-api-key"
npm run devFor production:
# Either with environment variable
export GHOST_SECURITY_API_KEY="your-api-key"
npm start
# Or with command line argument
npm run build
node dist/index.js "your-api-key"CLI Chatbot
This project includes a simple CLI chatbot powered by Anthropic's Claude API that can interact with the MCP server.
export ANTHROPIC_API_KEY="your-anthropic-api-key"
export GHOST_SECURITY_API_KEY="your-ghost-api-key"
npm run chatAdding to Any MCP Client
This server follows the standard MCP configuration format. Here are examples for different clients:
Standard MCP Configuration (Claude Desktop, ChatGPT, etc.)
{
"mcpServers": {
"ghostsecurity": {
"command": "node",
"args": ["/path/to/ghost-mcp-server/dist/index.js"],
"env": {
"GHOST_SECURITY_API_KEY": "your-ghost-security-api-key",
"GHOST_SECURITY_REPO_ID": "your-repository-id"
}
}
}
}Environment Variable Setup (Universal)
For any MCP client that supports environment variables:
export GHOST_SECURITY_API_KEY="your-ghost-security-api-key"
export GHOST_SECURITY_BASE_URL="https://api.ghostsecurity.ai/v2"
export GHOST_SECURITY_REPO_ID="your-repository-id"Repository Scoping
When a repository ID is configured via GHOST_SECURITY_REPO_ID environment variable or CLI argument:
ghostsecurity_get_findingsreturns findings only for that repositoryghostsecurity_get_repositoriesreturns only the configured repositoryghostsecurity_get_repository_findingsuses the configured repository as default (repoId parameter becomes optional)
Available Tools
Security Findings
ghostsecurity_get_findings
Get security findings with optional filtering and pagination.
Parameters:
cursor(string, optional): Pagination cursorsort(string, optional): Sort field (created_at,updated_at)order(string, optional): Sort order (asc,desc)size(number, optional): Page size (1-1000)status(string, optional): Filter by statusrepo_id(string, optional): Filter by repository IDproject_id(string, optional): Filter by project ID
ghostsecurity_get_finding
Get a specific security finding by ID.
Parameters:
id(string, required): Finding IDrepoId(string, required): Repository ID associated with the findingprojectId(string, required): Project ID associated with the finding
ghostsecurity_update_finding_status
Update the status of a security finding.
Parameters:
id(string, required): Finding IDrepoId(string, required): Repository ID associated with the findingprojectId(string, required): Project ID associated with the findingstatus(string, required): New status for the finding
Repositories
ghostsecurity_get_repositories
Get repositories with optional filtering and pagination.
Parameters:
cursor(string, optional): Pagination cursorsort(string, optional): Sort field (created_at,updated_at,last_committed_at)order(string, optional): Sort order (asc,desc)size(number, optional): Page size (1-1000)
ghostsecurity_get_repository
Get a specific repository by ID.
Parameters:
id(string, required): Repository ID
ghostsecurity_get_repository_findings
Get security findings for a specific repository.
Parameters:
repoId(string, optional*): Repository ID (*required unlessGHOST_SECURITY_REPO_IDis configured)cursor(string, optional): Pagination cursorsort(string, optional): Sort field (created_at,updated_at)order(string, optional): Sort order (asc,desc)size(number, optional): Page size (1-1000)
Publishing to MCP Registry
This server is configured for publishing to the MCP Registry.
Install Publisher CLI:
brew install mcp-publisherLogin:
mcp-publisher login githubPublish:
mcp-publisher publish
The server.json file contains the necessary metadata. Ensure you have published the package to NPM before publishing to the registry.
Development
Scripts
npm run dev: Run in development mode with hot reloadnpm run build: Build TypeScript to JavaScriptnpm start: Run the built servernpm run chat: Run the CLI chatbot
Project Structure
src/
├── index.ts # Main MCP server implementation
├── ghost-client.ts # Ghost Security API client (V2)
├── types.ts # TypeScript type definitions
└── chatbot.ts # CLI Chatbot implementationAPI Reference
This MCP server provides a complete interface to the Ghost Security API V2. For detailed API documentation, visit: https://docs.ghostsecurity.ai/api-reference/introduction
License
MIT
Available Tools
7 toolsghostsecurity_count_findingsB
Get count and statistics of security findings
Inputs: sort, order, status, repo_id, +1 more.
ghostsecurity_get_findingA
Get a specific security finding by ID (requires repository and project context)
Inputs: id (required), repoId (required), projectId (required).
ghostsecurity_get_findingsB
Get security findings with optional filtering and pagination
Inputs: cursor, sort, order, size, +5 more.
ghostsecurity_get_repositoriesB
Get repositories with optional filtering and pagination
Inputs: cursor, sort, order, size.
ghostsecurity_get_repositoryA
Get a specific repository by ID
Inputs: id (required).
ghostsecurity_get_repository_findingsC
Get security findings for a specific repository
Inputs: repoId (required), cursor, sort, order, +4 more.
ghostsecurity_update_findingA
Update a security finding (status and/or comments). At least one of status or comments must be provided.
Inputs: id (required), repoId (required), projectId (required), status, +1 more.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 7 tools
Each tool targets a distinct operation: counting, retrieving specific findings, filtering findings, retrieving repositories, retrieving a specific repository, repository-scoped findings, and updating findings. No overlapping purposes exist.
All tools follow a consistent 'ghostsecurity_verb_noun' pattern (e.g., ghostsecurity_get_findings, ghostsecurity_update_finding). The naming is predictable and uniform.
7 tools is well-scoped for a security findings server, covering essential read and limited write operations without being excessive or insufficient.
The tool surface covers core functionality (CRUD for findings, read for repositories) but lacks operations like creating or deleting findings, which may be intentional given typical security workflows.
Maintenance
Related MCP Connectors
MCP server for Pentest-Tools.com: run scans, manage findings and reports via your preffered LLM.
MCP server for the Vulners.com API
MCP server for mandates, delegation, policy-gated execution, credential grants, and audit.
Security scanner for MCP servers. Detect vulnerabilities, prompt injection, and tool poisoning.
Related MCP Servers
- AlicenseAqualityDmaintenanceA Model Context Protocol server that facilitates integration with OpenCTI, allowing users to query and retrieve cyber threat intelligence data via a standardized interface.1643MIT
- AlicenseCqualityDmaintenanceA Model Context Protocol server designed for testing backend APIs for security vulnerabilities like authentication bypass, injection attacks, and data leakage.1416MIT
- AlicenseAqualityAmaintenanceA Model Context Protocol (MCP) server for Socket integration, allowing AI assistants to efficiently check dependency vulnerability scores and security information.14,996 npm142MIT
- AlicenseNot gradedqualityCmaintenanceModel Context Protocol server for security research automation, integrating multiple security testing tools into LLM-driven workflows for secret scanning, static analysis, and vulnerability discovery.57Apache 2.0