FortiCNAPP MCP Server
MCP server for managing Fortinet FortiCNAPP via AI agents, enabling vulnerability scanning, agent management, and compliance monitoring.
README
FortiCNAPP MCP Server
MCP server for managing Fortinet FortiCNAPP (Cloud Native Application Protection Platform) via AI agents. Built with FastMCP, deployed as a container.
FortiCNAPP (powered by Lacework) provides cloud security capabilities including vulnerability scanning, agent management, and compliance monitoring.
Tools
| Tool | Description |
|---|---|
cnapp_health_check |
Check FortiCNAPP service health and connectivity |
cnapp_validate_config |
Validate configuration settings and credentials |
cnapp_get_agent_tokens |
Retrieve agent access tokens |
cnapp_scan_image_vulnerabilities |
Scan container images for vulnerabilities |
Every tool accepts optional cnapp_key_id, cnapp_key_secret, and cnapp_base_url parameters. If not provided, the server reads from environment variables. Per-call parameters override environment variables.
Connect from Claude Desktop
Add to your Claude Desktop config (~/Library/Application Support/Claude/claude_desktop_config.json on macOS):
{
"mcpServers": {
"forticnapp": {
"command": "npx",
"args": [
"-y",
"mcp-remote",
"https://mcp-forticnapp.fortidemoscloud.com/mcp"
]
}
}
}
Connect from Gemini CLI
Add to your Gemini settings (~/.gemini/settings.json):
{
"mcpServers": {
"forticnapp": {
"command": "npx",
"args": [
"-y",
"mcp-remote",
"https://mcp-forticnapp.fortidemoscloud.com/mcp"
]
}
}
}
Connect from Kiro / VS Code
Add to .kiro/settings/mcp.json or equivalent:
{
"mcpServers": {
"forticnapp": {
"url": "https://mcp-forticnapp.fortidemoscloud.com/mcp"
}
}
}
Test with curl
# 1. Initialize session and capture Mcp-Session-Id from headers
export SESSION_ID=$(curl -s -i -X POST https://mcp-forticnapp.fortidemoscloud.com/mcp \
-H "Content-Type: application/json" \
-d '{"jsonrpc":"2.0","id":1,"method":"initialize","params":{"protocolVersion":"2024-11-05","capabilities":{},"clientInfo":{"name":"test-curl","version":"1.0"}}}' \
| grep -i "mcp-session-id" | awk '{print $2}' | tr -d '\r')
echo "Session ID: $SESSION_ID"
# 2. List tools using the captured Session ID
curl -s -X POST https://mcp-forticnapp.fortidemoscloud.com/mcp \
-H "Content-Type: application/json" \
-H "Mcp-Session-Id: $SESSION_ID" \
-d '{"jsonrpc":"2.0","id":2,"method":"tools/list","params":{}}'
# 3. Call a tool (health check)
curl -s -X POST https://mcp-forticnapp.fortidemoscloud.com/mcp \
-H "Content-Type: application/json" \
-H "Mcp-Session-Id: $SESSION_ID" \
-d '{"jsonrpc":"2.0","id":3,"method":"tools/call","params":{"name":"cnapp_health_check","arguments":{"cnapp_key_id":"YOUR_KEY_ID","cnapp_key_secret":"YOUR_KEY_SECRET","cnapp_base_url":"https://youraccount.lacework.net"}}}'
# 4. Scan image vulnerabilities
curl -s -X POST https://mcp-forticnapp.fortidemoscloud.com/mcp \
-H "Content-Type: application/json" \
-H "Mcp-Session-Id: $SESSION_ID" \
-d '{"jsonrpc":"2.0","id":4,"method":"tools/call","params":{"name":"cnapp_scan_image_vulnerabilities","arguments":{"image_digest":"sha256:abc123...","cnapp_key_id":"YOUR_KEY_ID","cnapp_key_secret":"YOUR_KEY_SECRET"}}}'
Run locally
# Docker (with credentials from environment)
export FORTICNAPP_KEY_ID="your_key_id"
export FORTICNAPP_KEY_SECRET="your_key_secret"
export FORTICNAPP_BASE_URL="youraccount.lacework.net"
docker-compose up --build -d
# Or directly
uv sync
FORTICNAPP_KEY_ID="your_key_id" \
FORTICNAPP_KEY_SECRET="your_key_secret" \
FORTICNAPP_BASE_URL="youraccount.lacework.net" \
uv run uvicorn app.server:app --host 0.0.0.0 --port 8000
Server available at http://localhost:8000/mcp with health check at /health.
Deploy to Kubernetes
kubectl apply -f k8s-deployment.yaml
Exposes on NodePort 30083. Image: jviguerasfortinet/mcp-forticnapp-server:v1.0.0
Environment Variables
| Variable | Required | Default | Description |
|---|---|---|---|
FORTICNAPP_KEY_ID |
Yes | — | Lacework API key ID |
FORTICNAPP_KEY_SECRET |
Yes | — | Lacework API key secret (X-LW-UAKS value) |
FORTICNAPP_BASE_URL |
No | lwintseemea-eu.lacework.net |
Lacework API base URL or FQDN. Can be a full URL (https://myaccount.lacework.net) or just the FQDN (myaccount.lacework.net) — https:// is auto-prepended if missing. |
Tool Parameters
cnapp_health_check / cnapp_validate_config / cnapp_get_agent_tokens
| Parameter | Required | Description |
|---|---|---|
cnapp_key_id |
No | FortiCNAPP API key ID (uses FORTICNAPP_KEY_ID env var if not provided) |
cnapp_key_secret |
No | FortiCNAPP API key secret (uses FORTICNAPP_KEY_SECRET env var if not provided) |
cnapp_base_url |
No | FortiCNAPP API base URL or FQDN (e.g., myaccount.lacework.net). https:// is auto-prepended if missing. Uses FORTICNAPP_BASE_URL env var if not provided. |
cnapp_scan_image_vulnerabilities
| Parameter | Required | Default | Description |
|---|---|---|---|
image_digest |
Yes | — | Docker image digest (e.g., sha256:abc123...) |
cnapp_key_id |
No | — | FortiCNAPP API key ID |
cnapp_key_secret |
No | — | FortiCNAPP API key secret |
cnapp_base_url |
No | — | FortiCNAPP API base URL or FQDN. https:// is auto-prepended if missing. |
days_back |
No | 3 | Number of days to look back for scan data |
deduplicate |
No | true | Remove duplicate vulnerabilities across layers |
Authentication
The server uses Lacework bearer token authentication:
- Token Generation: Uses
FORTICNAPP_KEY_SECRET(X-LW-UAKS header) andFORTICNAPP_KEY_IDto request bearer tokens from/api/v2/access/tokens - Token Caching: Automatically caches tokens and refreshes before expiration (with 60s buffer)
- API Calls: All Lacework API calls use the cached bearer token in the
Authorization: Bearer <token>header - Retry Logic: Automatic retry with exponential backoff for transient failures
- URL Normalization:
FORTICNAPP_BASE_URLaccepts either a full URL (https://myaccount.lacework.net) or just the FQDN (myaccount.lacework.net) — thehttps://scheme is always auto-prepended if missing
License
MIT
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。