CAMEL AI Blog Translator Agent with Mistral LLM

import os
from camel.models import ModelFactory
from camel.types import ModelPlatformType, ModelType
from camel.messages import BaseMessage
from camel.agents import ChatAgent

# Ensure the MISTRAL_API_KEY environment variable is set
if not os.getenv('MISTRAL_API_KEY'):
  raise EnvironmentError("MISTRAL_API_KEY environment variable is not set.")

# Create the Mistral AI model using the Mistral platform and model type.
# The temperature is set to 0.0 for deterministic responses (no randomness in output).
mistral_model = ModelFactory.create(
  model_platform=ModelPlatformType.MISTRAL,
  model_type=ModelType.MISTRAL_LARGE,
  model_config_dict={"temperature": 0.0},
)

# The assistant is tasked with translating blog posts from English to German (Germany).
# It should only output the translation in MDX format without extra comments.
german_system_message = BaseMessage.make_assistant_message(
  role_name="translator",
  content="""
      You are a specialized translator of blog posts written in English to German (Germany).
      You will receive an English post in MDX format.
      Translate to German using technical language and do not translate English technical terms.
      Very important: Only write in the output the translated post in MDX format inside a snippet without other comments in the format of a snippet exactly like:
      ```mdx
      <mdx_contents>
      ```
  """
)

# Create the ChatAgent with the Mistral model and the specified system message.
# This agent will handle translation requests according to the defined role and instructions.
german_agent = ChatAgent(model=mistral_model, system_message=german_system_message)

# Define the content for the user message in MDX format.
# This content represents a blog post to be translated by the AI Agent.
content = """
---
title: CAMEL AI Blog Translator Agent with Mistral LLM  
description: "A post on using CAMEL AI Agents with Mistral LLM for a practical blog translation use case."  
date: Oct 09 2024  
lastUpdate: Oct 09 2024  
tags: ["CAMEL AI", "MDX", "Mistral AI", "agent", "blog post", "german", "german_agent", "german_blog post", "german_post", "user", "user_message"]
category: "AI Agents"
---

While it's easy to recognize the immense potential of AI Agents, it's not always straightforward to find a useful application for them in professional or personal projects.

In this post, I describe using the CAMEL AI framework to create a very simple AI Agent powered by Mistral LLMs, designed to assist with my blog project.

The goals of this post are twofold:
1. To describe a very simple AI Agent using the CAMEL AI framework and Mistral AI LLM API endpoint.
2. To provide a practical use case for this AI Agent.

Link(s):  
[CAMEL AI](https://docs.camel-ai.org)  
[Mistral AI API](https://docs.mistral.ai/api/) 
"""

# Create a user message with the role 'user' and the blog post content.
# This message is sent to the AI Agent for translation.
user_message = BaseMessage.make_user_message(role_name="user", content=content)

# Invoke the AI Agent with the user message and retrieve the agent's response.
# The response should contain the translated blog post in MDX format.
agent_response = german_agent.step(user_message)

# Output the translated content from the agent's response.
print(agent_response.msgs[0].content)