Interface Agent

User Interaction Agent is the main interface for receiving user instructions, coordinating multi-agent tasks, and logging conversations via the terminal.

Responsibility

User Interaction Agent acts as the main interface for coordinating user instructions and managing multi-agent tasks. It interacts with the user via terminal and orchestrates requests among various agents, ensuring seamless workflow and conversation logging.

Details

• Framework: LangChain
• Tools used: Coral MCP Tools, ask_human Tool (human-in-the-loop)
• AI model: GPT-4.1, GROQ-llama-3.3-70b
• Date added: June 4, 2025
• License: MIT

Setup the Agent

1. Clone & Install Dependencies

Details

# In a new terminal clone the repository:
git clone https://github.com/Coral-Protocol/Coral-Interface-Agent.git

# Navigate to the project directory:
cd Coral-Interface-Agent

# Download and run the UV installer, setting the installation directory to the current one
curl -LsSf https://astral.sh/uv/install.sh | env UV_INSTALL_DIR=$(pwd) sh

# Create a virtual environment named `.venv` using UV
uv venv .venv

# Activate the virtual environment
source .venv/bin/activate

# install uv
pip install uv

# Install dependencies from `pyproject.toml` using `uv`:
uv sync

2. Configure Environment Variables

Get the API Key: OpenAI or GROQ

Details

# Create .env file in project root
cp -r .env_sample .env

Run the Agent

You can run in either of the below modes to get your system running.

• The Executable Model is part of the Coral Protocol Orchestrator which works with Coral Studio UI.
• The Dev Mode allows the Coral Server and all agents to be seperately running on each terminal without UI support.

1. Executable Mode

Checkout: How to Build a Multi-Agent System with Awesome Open Source Agents using Coral Protocol and update the file: coral-server/src/main/resources/application.yaml with the details below, then run the Coral Server and Coral Studio UI. You do not need to set up the .env in the project directory for running in this mode; it will be captured through the variables below.

Details

For Linux or MAC:

registry:
  # ... your other agents
  interface-agent:
    options:
      - name: "MODEL_API_KEY"
        type: "string"
        description: "API key for the model provider"
      - name: "MODEL_NAME"
        type: "string"
        description: "What model to use (e.g 'gpt-4.1')"
        default: "gpt-4.1"
      - name: "MODEL_PROVIDER"
        type: "string"
        description: "What model provider to use (e.g 'openai', etc)"
        default: "openai"
      - name: "MODEL_MAX_TOKENS"
        type: "string"
        description: "Max tokens to use"
        default: 16000
      - name: "MODEL_TEMPERATURE"
        type: "string"
        description: "What model temperature to use"
        default: "0.3"

    runtime:
      type: "executable"
      command: ["bash", "-c", "<replace with path to this agent>/run_agent.sh main.py"]
      environment:
        - option: "MODEL_API_KEY"
        - option: "MODEL_NAME"
        - option: "MODEL_PROVIDER"
        - option: "MODEL_MAX_TOKENS"
        - option: "MODEL_TEMPERATURE"

For Windows, create a powershell command (run_agent.ps1) and run:

command: ["powershell","-ExecutionPolicy", "Bypass", "-File", "${PROJECT_DIR}/run_agent.ps1","main.py"]

2. Dev Mode

Ensure that the Coral Server is running on your system and run below command in a separate terminal.

Details

# Run the agent using `uv`:
uv run python main.py

You can view the agents running in Dev Mode using the Coral Studio UI by running it separately in a new terminal.

Example

Details

# Input:
Agent: How can I assist you today?

#Output:
The agent will interact with you directly in the console and coordinate with other agents as needed.

Creator Details

• Name: Suman Deb
• Affiliation: Coral Protocol
• Contact: Discord