A command-line interface (CLI) that uses the Gemini API to generate news articles from a text file or a URL.

This project uses the Google Gemini API. You need to provide your API key in an environment file.

1. Create a file named .env in the root of the project.

2. Add your API key to the file like this:

   GOOGLE_API_KEY="YOUR_API_KEY_HERE"
   

Note: The .env file is included in .gitignore and should not be committed to your repository.

You can configure a default input file using an environment variable:

# In your .env file
NEWS_INPUT_FILE="/path/to/your/default/file.txt"

Or set it in your shell:

export NEWS_INPUT_FILE="/path/to/your/default/file.txt"

You can configure a default output directory for saving generated files using an environment variable:

# In your .env file
NEWS_OUTPUT_DIR="/path/to/your/output/directory"

Or set it in your shell:

export NEWS_OUTPUT_DIR="/path/to/your/output/directory"

When this variable is set, the generated news and Bluesky files will be automatically saved to this directory without needing to specify --output-dir each time.

First, clone the repository and navigate into the project directory:

git clone https://github.com/fernand0/news_manager.git
cd news_manager

It is highly recommended to use a virtual environment.

Using venv:

python -m venv .venv
source .venv/bin/activate

Using uv:

uv venv
source .venv/bin/activate

Install the project in "editable" mode, which also installs all the necessary dependencies defined in pyproject.toml.

Using pip:

pip install -e .

Using uv:

uv pip install -e .

Generates a news story by sending the content of an input file or a URL to the Gemini API.

Input options (exclusive):

• --input-file / -i: Local text file
• --url: URL of an online news article

Customization options:

• --prompt-extra: Additional instructions to customize the generation
• --interactive-prompt: Interactive mode to add custom instructions
• --output-dir: Directory to save the generated files (default: uses NEWS_OUTPUT_DIR or does not save files)

Note: You cannot use both input options at the same time. If you use both, the command will show an error.

Usage examples:

1. From a local file

   python -m news_manager generate --input-file ./my_news.txt
   # or
   python -m news_manager generate -i ./my_news.txt

2. From a URL

   python -m news_manager generate --url "https://www.bbc.com/news/world-xxxxxx"

3. Default file

   python -m news_manager generate
   # It will use /tmp/noticia.txt or the file configured in NEWS_INPUT_FILE

4. With custom instructions

   # Focus on a specific person
   python -m news_manager generate --url "https://example.com/news" --prompt-extra "focus on María Santos and ignore the rest"
   
   # Focus on a particular aspect
   python -m news_manager generate -i news.txt --prompt-extra "focus only on the technological aspects"
   
   # Change the tone
   python -m news_manager generate --url "https://example.com" --prompt-extra "use a more formal and academic tone"
   
   # Interactive mode (you will be asked for instructions)
   python -m news_manager generate --url "https://example.com" --interactive-prompt

5. Saving generated files

   # Specify output directory manually
   python -m news_manager generate --url "https://example.com" --output-dir ./my_news
   
   # Use directory configured in NEWS_OUTPUT_DIR
   export NEWS_OUTPUT_DIR="/home/user/news"
   python -m news_manager generate --url "https://example.com"
   # Files will be automatically saved in /home/user/news

When the generated news corresponds to a doctoral thesis, the system produces:

• Title: In the form PhD Thesis Defense of [name] [first surname], "[thesis title]".
• Slug/URL: The generated file will include the author's name and first surname, followed by keywords from the thesis title.

Example:

• Generated title:

  PhD Thesis Defense of Name Surname, "The Title"
  

• Generated file:

  2025-07-15-name-surname-the-title.txt
  

This facilitates the identification and access to thesis news, and improves the readability of files and URLs.

Advanced options:

# See command help
python -m news_manager generate --help

Example Output:

The output will be the formatted news article generated by the Gemini API.

--- Initializing AI client and generating news... ---
Title: New species of luminescent butterfly discovered in the Amazon by doctors Alistair Finch and Elena Vance
Text: Dr. Alistair Finch, a biologist from Greendale University, along with Dr. Elena Vance, has announced the discovery of a new species of butterfly that glows in the dark in the Amazon rainforest. This finding has been published in a recent scientific study.

The research details the unique characteristics of this insect, which uses bioluminescence as a possible defense and communication mechanism. Dr. Finch has dedicated more than a decade to the study of entomology in the region, while Dr. Vance is an expert in insect genetics. Greendale University has financed a large part of this expedition.
Links:
- https://example.com/original-news
Bluesky: A new species of luminescent butterfly has been discovered in the Amazon by doctors Alistair Finch and Elena Vance. This incredible finding opens new doors for bioluminescence research. #science #nature #discovery [link to news]

Prints a welcome message to confirm that the news-manager CLI is correctly installed and accessible in your environment.

How to run:

python -m news_manager hello

Example Output:

Welcome to News Manager CLI!
This tool helps you generate and manage news content using the Gemini API.
Try 'python -m news_manager generate' to create a news story or 'python -m news_manager --help' for more commands.

This project was developed using a collaborative approach with AI tools:

• Google Gemini AI: Used for the AI-powered news generation functionality and initial development
• Cursor: The world's best AI-powered code editor was used for pair programming, code generation, and iterative development

The development process involved a conversational flow to build, refactor, and enhance the application step by step, leveraging the strengths of both tools for optimal code quality and functionality.

Visit our beautiful website to learn more about News Manager:

📰 News Manager Website

The website features:

• ✨ Modern, responsive design
• 📱 Mobile-friendly interface
• 🚀 Interactive examples
• 📊 Live demonstrations
• 🛠️ Complete documentation

This project is licensed under the MIT License - see the LICENSE file for details.

The project includes an auxiliary CLI to publish posts to Bluesky from the files generated by News Manager.

• news_manager/: Main package for news generation and main CLI.
• news_publisher/: Auxiliary tools for publishing to Bluesky (own CLI).

1. Post search directory

   • By default, the CLI searches for *_blsky.txt files in the directory configured in the BLUESKY_POSTS_DIR variable of your .env:

     BLUESKY_POSTS_DIR=/path/to/your/directory
     

   • If not defined, it will use the current directory.
   • You can override it with the --dir argument.

2. Bluesky User

   • By default, the last profile defined in ~/.mySocial/config/.rssBlsk is used.
   • You can override it with the --user argument.

To publish the last generated post to Bluesky, you can use uv run or execute the module directly:

Using uv run:

uv run news_publisher publish

Executing the module:

python -m news_publisher publish

Options:

• --dir /path/to/files to specify the search directory.
• --user your_bluesky_user to specify the Bluesky user.

Example:

uv run news_publisher publish --dir /home/user/news --user myuser

The CLI will show the content to be published and ask for confirmation before sending it to Bluesky.

• Have the .env file configured with the BLUESKY_POSTS_DIR variable if you want a default directory.
• Have the ~/.mySocial/config/.rssBlsk file configured with your Bluesky credentials.
• Dependencies are managed automatically from pyproject.toml.

For specific conventions on the format of titles, slugs, and special rules (for example, for doctoral theses), see the CONVENTIONES.md file.