Skip to content

kleysonfiretail/firetail-mcp-server

BranchesTags

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

6 Commits
src
src
 
 
.gitignore
.gitignore
 
 
README.md
README.md
 
 
package-lock.json
package-lock.json
 
 
package.json
package.json
 
 
tsconfig.json
tsconfig.json
 
 

Repository files navigation

FireTail Resource Policy MCP Server

A Model Context Protocol (MCP) server that provides Claude Desktop with direct access to the FireTail Resource Policy API. This enables AI-assisted management of resource policies, filters, and alert configurations through natural language.

Features

  • 🔐 OAuth2 Authentication - Automatic token management with client credentials flow
  • 🔄 Auto Token Refresh - Handles token expiration seamlessly
  • 🛠️ 8 Core Tools - Full CRUD operations for resource policies
  • 📝 TypeScript - Full type safety with comprehensive type definitions
  • 🌍 Multi-Region - Support for EU and US endpoints
  • ⚡ Production Ready - Error handling, retry logic, and logging

Tools Available

Tool Description
list_resource_policies List all resource policies for an organisation
list_premade_filters List all premade filters for an organisation
list_resource_items List all resource items for an organisation
list_existing_integrations_for_policy List all existing notification integrations for an organisation
create_resource_policy Create a new resource policy
get_resource_policy Get a resource policy by UUID
update_resource_policy Update an existing resource policy
delete_resource_policy Delete a resource policy

Prerequisites

  • Node.js 18+ (with native fetch support)
  • Claude Desktop application
  • FireTail OAuth2 credentials (client ID and secret)
  • Your FireTail organisation UUID

1. Clone and Install

# Clone repo
git clone https://github.com/kleysonfiretail/firetail-mcp-server.git
cd firetail-mcp-server

# Install dependencies
npm i

2. Build the Project

npm run build

3 Get FireTail Credentials

Obtain your OAuth2 credentials from FireTail:

  • CLIENT_ID: Your OAuth2 client ID
  • CLIENT_SECRET: Your OAuth2 client secret

Test your credentials:

curl -X POST https://api.saas.eu-west-1.prod.firetail.app \
  -d grant_type=client_credentials \
  -d client_id=YOUR_CLIENT_ID \
  -d client_secret=YOUR_CLIENT_SECRET

4. Configure Claude Desktop

Edit Claude Desktop's config file:

MacOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json

Add this configuration:

{
  "mcpServers": {
    "firetail-resource-policy": {
      "command": "node",
      "args": ["/absolute/path/to/firetail-mcp-server/dist/index.js"],
      "env": {
        "FIRETAIL_API_URL": "https://api.saas.eu-west-1.prod.firetail.app",
        "FIRETAIL_CLIENT_ID": "your-client-id-here",
        "FIRETAIL_CLIENT_SECRET": "your-client-secret-here",
        "FIRETAIL_DEFAULT_ORG_UUID": "your-org-uuid", 
        "FIRETAIL_DEFAULT_PLATFORM_TYPE": "ai" 
      }
    }
  }
}

Important: Use the absolute path to your dist/index.js file!

Region Configuration

EU (default):

"FIRETAIL_API_URL": "https://api.saas.eu-west-1.prod.firetail.app"

US:

"FIRETAIL_API_URL": "https://api.saas.us-east-2.prod.us.firetail.app"

5. Restart Claude Desktop

Completely quit and restart Claude Desktop for the configuration to take effect.

6. Usage Examples

Once configured, you can use natural language to interact with the FireTail API:

1. Listing & Searching

List all resource policies

"List all resource policies"
"Show me all policies"

Search for specific policies

"List all resource policies with 'deepseek' in the name"
"Find policies named with 'test'"
"Search for policies containing 'production'"

List suggested policies (premade filters)

"List all suggested policies"
"Show me available premade filters"
"What suggested policies are available for AI platform?"

List notification integrations

"List all existing integrations"
"Show me Slack integrations"
"Find integrations named with 'email'"

List resource items

"List all resource items for AI platform"
"Show me available resources"

2. Viewing Policy Details

Get specific policy information

"Show me details of policy UUID abc-123"
"Get the resource policy with UUID xyz-789"
"Load policy abc-123 with customizations"

3. Creating Policies

Basic policy creation

"Create a new resource policy called 'Production Monitor' with description 'Monitor prod AI usage'"
"Create a policy named 'DeepSeek Alert' using the DeepSeek suggested policy"

Policy with notifications

"Create a policy called 'Critical Alerts' using suggested policy 1 and send notifications to Slack UUID abc-123"
"Create 'Team Notifications' policy with multiple Slack channels"

Policy with custom filters

"Create a policy with custom filters for specific LLM models"
"Create a policy that monitors OpenAI GPT-4 usage in GitHub"

4. Updating Policies

Update policy name

"Update the resource policy named 'testing' to 'Claude MCP'"
"Rename policy 'old-name' to 'new-name'"

Update policy description

"Update the description of policy 'Monitor AI' to 'Track all AI model usage'"
"Change the description for policy UUID abc-123"

Add/remove notifications

"Add Slack notification UUID xyz-789 to policy 'Production Monitor'"
"Remove notification from policy 'Test Policy'"

Modify filters

"Update policy 'DeepSeek Monitor' to include GitHub and GitLab platforms"
"Add custom filter for GPT-4 models to policy 'AI Tracking'"

5. Deleting Policies

Delete by name or UUID

"Delete the resource policy 'Show and Tell'"
"Delete policy with UUID abc-123"
"Remove the 'testing' policy"

6. Complex Workflows

Creating comprehensive monitoring

"Create a policy that monitors all DeepSeek usage in code repositories and sends alerts to the security team Slack channel"

"Set up a policy to track any AI model usage on cloud platforms, excluding GitHub, and notify via email"

Audit and cleanup

"List all policies created by riley@firetail.io"
"Find all disabled policies"
"Show me policies that haven't been updated in the last 30 days"

Bulk operations

"Show me all test policies so I can decide which to delete"
"List all policies using the 'Any AI Usage' suggested filter"

7. Integration Management

Finding the right integration

"List all Slack integrations for team notifications"
"Find the Slack channel integration for @user"
"Show me all enabled notification integrations"

Setting up alerts

"Create a policy using 'Any AI Usage' and send alerts to all team Slack channels"
"Set up monitoring for DeepSeek with notifications to security@company.com"

8. Policy Analysis

Understanding configurations

"What filters are applied to policy 'Production Monitor'?"
"Show me which notifications are configured for 'DeepSeek Alert'"
"What suggested policies are being used by 'AI Tracking'?"

Compliance and reporting

"List all policies monitoring AI usage in code"
"Show me policies that alert on high-risk models"
"Which policies are monitoring platform usage vs code usage?"

Project Structure

firetail-mcp-server/
├── src/
│   └── index.ts           # Main MCP server code
├── dist/                  # Compiled JavaScript (generated)
│   └── index.js
├── package.json           # Project configuration
├── tsconfig.json          # TypeScript configuration
└── README.md             # This file

Development

Run in Development Mode

npm run dev

Type Checking

npx tsc --noEmit

Build for Production

npm run build

View Logs

MacOS:

tail -f ~/Library/Logs/Claude/mcp*.log

Windows:

Get-Content "$env:APPDATA\Claude\logs\mcp*.log" -Wait

Troubleshooting

Server Not Appearing

Check:

  • Configuration file is valid JSON
  • Absolute path is correct
  • Claude Desktop was fully restarted
  • Build succeeded (dist/index.js exists)

Authentication Errors

Verify:

  • CLIENT_ID and CLIENT_SECRET are correct
  • Credentials haven't expired
  • You have access to the organisation
  • API URL matches your region

Test manually:

curl -X POST https://api.saas.eu-west-1.prod.firetail.app \
  -d grant_type=client_credentials \
  -d client_id=YOUR_CLIENT_ID \
  -d client_secret=YOUR_CLIENT_SECRET

Security Best Practices

✅ Never commit credentials to version control
✅ Use environment variables for sensitive data
✅ Rotate credentials regularly
✅ Limit credential permissions to required scope
✅ Monitor API usage through FireTail dashboard
✅ Keep dependencies updated for security patches

Using Environment Variables

Create a .env file (don't commit this):

FIRETAIL_API_URL=https://api.saas.eu-west-1.prod.firetail.app
FIRETAIL_CLIENT_ID=your-client-id
FIRETAIL_CLIENT_SECRET=your-client-secret
FIRETAIL_DEFAULT_ORG_UUID=your-org-uuid
FIRETAIL_DEFAULT_PLATFORM_TYPE=ai

Endpoints

All API endpoints are documented in the FireTail OpenAPI specification. The server provides MCP tools that map to these endpoints:

  • /organisations/{org_uuid}/resource-policies/filters
  • /organisations/{org_uuid}/resource-policies/resources
  • /organisations/{org_uuid}/resource-policies
  • /organisations/{org_uuid}/resource-policies/{resource_policy_uuid}
  • /organisations/{org_uuid}/search

Support

For issues with:

Acknowledgments

Built with:

About

MCP Server for FireTail Resource Policy API (TypeScript)

www.firetail.app/

Resources

Readme
Activity

Stars

0 stars

Watchers

0 watching

Forks

0 forks
Report repository

Contributors

Languages