Walee's Blog

A blog to discuss all things Ai Agent Frameworks and Platforms

---

Create a Zerepy Agent that uses llama3.2:1b via Ollama

---

Used the following versions when writing this blog
python - 3.12.8 (had issues with 3.13.1)
poetry - 1.8.5

Check out my GitHub repository for the source code: Zerepy Hello World

Faster Route

git clone --branch ollama https://github.com/wale-e/ZerePy.git   # clone repository
cd zerepy  # change to the zerepy direction
# install python 3.12.8  (instructions below)
# install poetry 1.8.5   (instructions below)
poetry env use python3.12    # ensure poetry the correct Python version
poetry install --no-root   # install dependencies
poetry shell    # activate the virtual environment"
# follow step from "Run the agent below"

Longer Detailed Route

Installation

First install python (version 3.12.8) if you haven’t, Python downloads

Make sure python is present in your path

python3 --version

should return Python 3.12.8

Install Poetry for dependency management if you haven’t already

Follow the steps here to use the official installation

Clone the repository

git clone https://github.com/blorm-network/ZerePy.git

Go to the zerepy directory

cd zerepy

Install dependencies

poetry install --no-root

This will create a virtual environment and install all required dependencies.

Ensure poetry uses the correct Python version

poetry env use python3.12

In src/connections/ folder create ollama_connection.py

touch ./src/connections/ollama_connection.py

Copy the content of this file from github into ollama_connection.py

ollama_connection.py

Edit the src/agent.py

Change the section below

# Extract Twitter config
twitter_config = next((config for config in agent_dict["config"] if config["name"] == "twitter"), None)
if not twitter_config:
    raise KeyError("Twitter configuration is required")

# TODO: These should probably live in the related task parameters
self.tweet_interval = twitter_config.get("tweet_interval", 900)
self.own_tweet_replies_count = twitter_config.get("own_tweet_replies_count", 2)

To

# Extract Twitter config
twitter_config = next((config for config in agent_dict["config"] if config["name"] == "twitter"), None)
if twitter_config:
    # Set Twitter-specific variables only if the config exists
    self.tweet_interval = twitter_config.get("tweet_interval", 900)
    self.own_tweet_replies_count = twitter_config.get("own_tweet_replies_count", 2)
else:
    logger.info("No Twitter configuration found. Skipping Twitter integration.")

Change the section below

if not self.username:
    raise ValueError("Twitter username is required")

To

if not self.username:
    logger.info("Twitter username not configured. Skipping Twitter-specific functionality.")

Edit the src/connection_manager.py

After

from src.connections.farcaster_connection import FarcasterConnection

Add

from src.connections.ollama_connection import OllamaConnection

After

elif class_name == "eternalai":
    return EternalAIConnection

Add

elif class_name == "ollama":
    return OllamaConnection

Create an ollama.json file in the agents folder

#assumes you are at the root of project (zerepy)
touch ./agents/ollama.json

Copy the copy of this file from github into agents/ollama.json

ollama.json

Edit agents/general.json to make Ollama the default agent on startup

{
  "default_agent": "ollama"
}

Activate the virtual environment

poetry shell

Run the agent

poetry run python main.py

At the prompt ZerePy-CLI (ollama) type ‘chat’

ZerePy-CLI (ollama) > chat

At the You prompt ask the chatbot any question

You: What is quantum computing

---

This completes your hello world Zerepy introduction
Congratulations you have successful run your first Zerepy Ai Agent

---

Project maintained by wale-e Hosted on GitHub Pages — Theme by mattgraham