# coreos-ops-assistant

**Repository Path**: mirrors_coreos/coreos-ops-assistant

## Basic Information

- **Project Name**: coreos-ops-assistant
- **Description**: An AI-powered assistant to help monitor CoreOS pipelines
- **Primary Language**: Unknown
- **License**: Not specified
- **Default Branch**: main
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2025-11-15
- **Last Updated**: 2025-12-27

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# CoreOS Operations Assistant

This is an AI-powered assistant used by the CoreOS team to make operations (such as
pipeline monitoring) easier. It works in Slack or Matrix.

The assistant listens for direct mentions in a chat channel in which it was invited.

## Setup

### Local Python Environment

1.  **Install dependencies:**
    ```
    source env/bin/activate  # if using a virtualenv
    pip install -r requirements.txt
    ```

2.  **Set environment variables:**

For Jenkins:

    *   `JENKINS_URL`: The URL of your Jenkins server.
    *   `JENKINS_TOKEN`: Your Jenkins token (often an API token).

If using Slack:

    *   `SLACK_BOT_TOKEN`: Your Slack bot token.
    *   `SLACK_APP_TOKEN`: Your Slack app token.

If using Matrix:

    *   `MATRIX_SERVER`: Your homeserver (i.e.  https://matrix.org)
    *   `MATRIX_ACCESS_TOKEN`: Your device access token
    *   `MATRIX_ROOM`: Your room FQDN (i.e. `#tmp-coreos-pipeline-assistant-testing:matrix.org`)

If using GEMINI:

    *   `GEMINI_API_KEY`: (OPTIONAL) Your Gemini API key.

If using OpenRouter:

    *   `OPENROUTER_API_KEY`: (OPTIONAL) Your OpenRouter access key

## Usage

1.  **Run the bot:**
    ```
    source env.sh            # tokens env vars
    source env/bin/activate  # if using a virtualenv
    python main.py
    ```

2.  **Invite the bot to your channel.**

3.  **Mention the bot in a thread of a Jenkins failure notification.** The bot will then reply with a summary of the failure.

### Running with Podman

1.  **Build the container image:**
    ```
    podman build -t coreos-ops-assistant .
    ```

2.  **Run the container:**
    You can pass the environment variables directly with the `-e` flag, or you can use a `.env` file.


    **Using the `-e` flag:**
    ```
    # First export the appropriate variables in your environment based
    # on what you are using (i.e. slack/matrix and gemini/openrouter)
    export SLACK_BOT_TOKEN='your_bot_token'
    export SLACK_APP_TOKEN='your_app_token'
    export MATRIX_SERVER='https://matrix.org'
    export MATRIX_ACCESS_TOKEN='your_app_token'
    export MATRIX_ROOM='your_room_fqdn'
    export JENKINS_URL='https://your.jenkins.url'
    export JENKINS_TOKEN='your_jenkins_token'
    export GEMINI_API_KEY='your_api_key'
    export OPENROUTER_API_KEY='your_api_key'

    # Now run `podman` and it will pull the values from the variables
    # already set in your environment.
    podman run -it --rm        \
      -e SLACK_BOT_TOKEN       \
      -e SLACK_APP_TOKEN       \
      -e MATRIX_SERVER         \
      -e MATRIX_ACCESS_TOKEN   \
      -e MATRIX_ROOM           \
      -e JENKINS_URL           \
      -e JENKINS_TOKEN         \
      -e GEMINI_API_KEY        \
      -e OPENROUTER_API_KEY    \
      coreos-ops-assistant
    ```

    **Using a `.env` file:**
    Create a file named `.env` with the following content:
    ```
    SLACK_BOT_TOKEN=your_bot_token
    SLACK_APP_TOKEN=your_app_token
    MATRIX_SERVER=https://matrix.org
    MATRIX_ACCESS_TOKEN=your_app_token
    MATRIX_ROOM=your_room_fqdn
    JENKINS_URL=https://your.jenkins.url
    JENKINS_TOKEN=your_jenkins_token
    GEMINI_API_KEY=your_api_key
    OPENROUTER_API_KEY=your_api_key
    ```
    Then run the container with the `--env-file` flag:
    ```
    podman run -it --rm --env-file .env coreos-ops-assistant
    ```

## Testing

To run the unit tests, run the following command from the root of the project:
```
python3 test.py
```