Skip to main content
Agents are the building blocks of your workflows in Agility. Each agent performs a specific task, from generating text with AI models to reading emails and sending messages to Discord.

Available Agents

Text Generator

Generate text using OpenAI GPT or Anthropic Claude models

Gmail Reader

Read and filter emails from your Gmail account

Gmail Sender

Send automated emails through your Gmail account

Discord Messenger

Send messages to Discord channels via webhooks

GitHub Reader

Monitor GitHub repositories for commits and changes

How Agents Work

Agents in Agility follow a consistent pattern:
  1. Configuration - Each agent requires specific credentials and settings
  2. Input Structure - Agents can access data from previous steps in the workflow
  3. Execution - Agents perform their designated task when the workflow runs
  4. Output Structure - Agents produce structured data that subsequent steps can use

Agent Testing

Every agent provides two testing options:
  • Test Agent - Tests the agent in isolation with your current configuration
  • Test Workflow - Runs the entire workflow from the beginning to see how data flows through the agent

Common Features

Input Field Insertion

Most agents support inserting fields from previous workflow steps. Click on available input fields to insert them into prompts, email bodies, or message content.

Configuration Persistence

Agent configurations are automatically saved and encrypted. Your API keys and credentials are securely stored and only accessible to your workflows.

Error Handling

Agents provide detailed error messages when configuration is incorrect or when API calls fail, making it easy to troubleshoot issues.

Next Steps

Explore the documentation for each agent to learn about:
  • Required credentials and authentication
  • Available configuration parameters
  • Output data structure
  • Real-world usage examples

Build docs developers (and LLMs) love