MCP Tools & Commands

Once Station is connected to Claude Desktop via MCP, you have access to powerful tools for agent management and system integration. This page shows you exactly what you can accomplish and provides practical examples.

Core Station MCP Tools

Agent Management Tools

call_agent - Execute Station Agents

Purpose: Run any Station agent with a specific task
What it does: Executes an agent and returns the complete result including tool usage, steps taken, and duration

Example Commands:

Run Station agent disk-monitor with the task: check disk usage and alert if any partition is over 80%

Execute my database monitoring agent to check connection pool status

Call agent deployment-pipeline to deploy the latest code to staging environment

create_agent - Create New Agents

Purpose: Create new Station agents with specific tools and configuration
What it does: Sets up a new agent with dotprompt configuration, tool assignments, and environment settings

Example Commands:

Create a new Station agent named "Security Scanner" that uses filesystem and network tools to scan for vulnerabilities

Create an agent for monitoring AWS resources using the AWS CLI tools, max 10 steps

Set up a database backup agent that can access PostgreSQL and S3 storage tools

list_agents - Browse Available Agents

Purpose: Show all agents across environments with their status and capabilities
What it does: Returns agent details including ID, name, environment, tools, and recent activity

Example Commands:

List all my Station agents

Show me agents in the production environment

What agents do I have for infrastructure monitoring?

get_agent_details - Agent Information

Purpose: Get detailed information about a specific agent
What it does: Returns complete agent configuration, tool assignments, execution history, and performance metrics

Example Commands:

Show me details for Station agent security-scanner

Get the configuration and recent runs for my deployment agent

What tools does agent filesystem-monitor have access to?

Tool & Environment Discovery

discover_tools - Find Available Tools

Purpose: Discover what MCP tools are available in your Station setup
What it does: Scans environments and returns available tools with descriptions and capabilities

Example Commands:

What MCP tools are available in my Station setup?

Show me all filesystem tools I can assign to agents

List database tools available for agent creation

list_tools - Browse Tools by Environment

Purpose: List tools filtered by environment or server
What it does: Shows detailed tool information including parameters, descriptions, and which agents use them

Example Commands:

List all tools in the production environment

Show me what tools are available from the AWS MCP server

What monitoring tools can I use in development?

list_environments - Environment Overview

Purpose: Show available Station environments
What it does: Returns environment details including tool counts, agent counts, and configuration status

Example Commands:

What Station environments do I have configured?

Show me the difference between my dev and prod environments

Practical Usage Patterns

Infrastructure Management

System Monitoring Setup:

Create a Station agent that monitors system resources every hour and alerts via Slack when CPU exceeds 90% or disk space is low

Log Analysis:

Run my log analysis agent to scan the last 24 hours of application logs for errors and generate a summary report

Database Health Checks:

Execute agent db-health-monitor to check database connection pools, slow queries, and disk usage across all production databases

Deployment & CI/CD

Automated Deployments:

Create a deployment agent that can pull latest code, run tests, and deploy to staging with rollback capability

Infrastructure as Code:

Run my Terraform agent to plan and apply infrastructure changes for the new microservice

Health Checks After Deployment:

Execute the health check agent to verify all services are responding correctly after the deployment

Security Operations

Vulnerability Scanning:

Create a security agent that scans container images for vulnerabilities and checks for exposed secrets

Compliance Auditing:

Run agent compliance-auditor to perform a full compliance audit of our AWS infrastructure against SOC2 requirements

Access Review:

Execute the access review agent to audit user permissions and flag any overprivileged accounts

Development Workflow

Code Quality Checks:

Create an agent that runs linting, security scanning, and test coverage analysis on pull requests

Database Management:

Run my database agent to create a backup before the schema migration and set up the new tables

Environment Synchronization:

Execute the sync agent to update development environment with anonymized production data

Advanced MCP Integration

Multi-Step Workflows

Station agents can handle complex multi-step workflows:

Create a deployment agent that:
1. Runs full test suite
2. Builds and scans container images
3. Updates staging environment
4. Runs integration tests
5. Promotes to production if tests pass
6. Sends Slack notifications at each step

Cross-Environment Operations

List agents in both development and production environments and show me which ones need to be synchronized

Create matching monitoring agents in dev and prod with environment-specific thresholds

Tool Integration

Agents can use multiple tools in combination:

Create an incident response agent that can:
- Read logs from multiple systems
- Query databases for error patterns  
- Update PagerDuty incidents
- Send status updates to Slack
- Generate post-mortem reports

Tool Assignment Examples

When creating agents, you can specify exactly which tools they need:

Filesystem & System Tools

Create an agent with file operations, directory listing, and system monitoring tools

Database Tools

Set up an agent with PostgreSQL, MySQL, and Redis tools for database operations

Cloud Provider Tools

Create an AWS infrastructure agent with EC2, S3, RDS, and CloudWatch access

Communication Tools

Build a notification agent that can use Slack, email, and PagerDuty integrations

Monitoring Agent Performance

Track your agents through Claude Desktop:

Show me execution metrics for agent terraform-analyzer over the last week

Which agents are using the most resources?

Get performance details for my most frequently run agents

Troubleshooting Through MCP

Debug agent issues directly through Claude Desktop:

Check the configuration and recent error logs for agent deployment-pipeline

List all failed agent runs in the last 24 hours

Show me which agents have tool assignment issues

Security & Audit

Station provides complete audit trails accessible via MCP:

Show me all agent executions in the production environment today

List which users have created or modified agents this week

Get audit logs for any agents that accessed sensitive database credentials

Next Steps

Remember: All Station MCP interactions provide rich context and detailed results. Ask follow-up questions to drill down into specifics, and Claude Desktop will help you manage your infrastructure through Station’s intelligent agents.