🏃How To Run

Step 1 (Clone Repo)

Using the following command create a local instance of this repo.

git clone [email protected]:fabelis/agent.git

Step 2 (Configuration)

To customize your agent edit your config.json file in the root directory. Here's an example of what it looks like below.

{
    // Model used for prompting
    "completion_provider": "anthropic",
    // Model to convert data to embeddings, options include local | openai
    "embedding_provider": "local",
    // Vector Database to store embeddings, currently supports local | mongodb
    "db": "local",
    // list your clients here 
    // [KEYS] indicate the client's name 
    // [VALUE] is the clients config (if no config is needed use true)
    "clients": {
        "cli": true
    }
}

What We Support below shows all integrated providers/dbs/clients.

What We Support

Step 3 (Setup .env)

Only fill in values that are necessary. This is dependent on your config.json file in Step 2.

All secrets are stored within the .env file. You can use .env.example as a reference.

Environment shows an explanation for each environment variable as a reference.

Environment

Step 4 (Create Character)

All character files MUST be located in the characters folder in the root directory of the project.

This section explains what the character file is. Check it out for assistance!

🏌️Character File

The program defaults to using the character located under character.example.json. If you would like to load a custom character file use the --character flag when launching. Here's an example:

cargo run -- --character fabelis.json

Step 5 (Run The Agent)

If no clients are used in the config.json then nothing will happen when you run the program.

As explained above you can run the program using the command below, but to load your custom character use the character flag.

cargo run

To see a simple example of running your agent check out the link below:

CLI Chatbot