MCP Tools Overview

Actify provides 8 powerful MCP tools that AI agents can use to execute code, manage resources, and monitor executions.

Available Tools

Tool	Purpose	Category
code_execute	Execute Node.js code on-demand	Execution
code_schedule	Schedule recurring code execution	Scheduling
code_cancel	Cancel running or scheduled jobs	Management
execution_status	Get execution status and logs	Monitoring
list_workflows	List recent executions and schedules	Management
get_credits	Check credit balance and usage	Credits
manage_secrets	Create, list, and delete secrets	Security
system_status	Get system load and capacity info	Monitoring

Tool Categories

Execution Tools

Execute code immediately or on a schedule:

• code_execute: Run code once, get results immediately
• code_schedule: Set up recurring jobs with cron expressions

Management Tools

Control and monitor your workflows:

• code_cancel: Stop running executions or remove schedules
• list_workflows: See all your executions and scheduled jobs
• execution_status: Check progress and retrieve logs

Resource Tools

Manage credits and secrets:

• get_credits: Monitor your usage and balance
• manage_secrets: Secure storage for API keys and tokens

System Tools

Check system health and capacity:

• system_status: Real-time system load and queue statistics

Common Workflows

Execute and Monitor

// 1. Execute code
const exec = await code_execute({ code: "...", runtime: "node18" });

// 2. Check status
const status = await execution_status({ executionId: exec.id });

// 3. Get logs if needed
const logs = await execution_status({
  executionId: exec.id,
  includeLogs: true
});

Schedule with Secrets

// 1. Create secret
await manage_secrets({
  action: "create",
  name: "API_KEY",
  value: "secret_value"
});

// 2. Schedule job using secret
await code_schedule({
  code: "console.log(process.env.API_KEY);",
  cronExpression: "0 9 * * *",
  secrets: ["API_KEY"],
  runtime: "node18"
});

Resource Management

// 1. Check credits before execution
const credits = await get_credits();

if (credits.balance > 100) {
  // 2. Execute code
  await code_execute({ code: "..." });

  // 3. List all workflows
  const workflows = await list_workflows({ limit: 10 });
}

Tool Response Patterns

All tools follow consistent response patterns:

Success Response

{
  "success": true,
  "data": { /* tool-specific data */ },
  "metadata": {
    "timestamp": "2025-01-08T10:30:00Z",
    "executionTime": 1234
  }
}

Error Response

{
  "success": false,
  "error": {
    "code": "INSUFFICIENT_CREDITS",
    "message": "Not enough credits to execute",
    "details": { "required": 100, "available": 50 }
  }
}

Rate Limits

• code_execute: 100 executions/hour per user
• code_schedule: 50 schedules active per user
• manage_secrets: 100 secrets per user
• API calls: 1000 requests/hour per user

Authentication

All tools require a Personal Access Token (PAT):

// Set in MCP client configuration
{
  "env": {
    "ACTIFY_PAT": "your_pat_token_here"
  }
}

Generate a PAT from your Actify dashboard or CLI:

actify auth create-token --name "My MCP Client"

Next Steps

• Explore individual tool documentation
• See practical examples
• Learn about error handling