Skip to content

Configuration🔗

Brag AI offers several configuration options to customize its behavior. This guide explains the different ways to configure Brag AI.

Environment Variables🔗

Brag AI supports the following environment variables:

Authentication Tokens🔗

Environment Variable Description
GITHUB_API_TOKEN GitHub API token for authentication (required for private repositories or to increase rate limits)
OPENAI_API_KEY OpenAI API key for using OpenAI models
ANTHROPIC_API_KEY Anthropic API key for using Claude models
GEMINI_API_KEY Google API key for using Gemini models

You can set these environment variables in your shell profile or use a .env file in your project directory.

Example .env file:

GITHUB_API_TOKEN=ghp_your_github_token
OPENAI_API_KEY=sk-your_openai_key

Environment Variable Precedence🔗

Environment variables take precedence over command-line arguments. For example, if you set GITHUB_API_TOKEN in your environment and also provide --github-api-token on the command line, the environment variable will be used.

Command-Line Options🔗

Brag AI has several command-line options that can be used to configure its behavior:

Repository Selection🔗

Option Description
owner/repo The GitHub repository to generate the brag document from (positional argument)
--user The GitHub username to generate the brag document for

Time Range🔗

Option Description
--from The start date to generate the brag document for (format: YYYY-MM-DD)
--to The end date to generate the brag document for (format: YYYY-MM-DD)
--limit The maximum number of commits to include in the brag document

Authentication🔗

Option Description
--github-api-token The GitHub API token to use for authentication

Output🔗

Option Description
--output The path to save the generated brag document
--overwrite If set, overwrites the output file if it already exists

AI Model Configuration🔗

Option Description
--model The name of the AI model to use for generating the brag document
--language The language to use for generating the brag document

Model Selection🔗

Brag AI supports a variety of AI models through PydanticAI. You can specify which model to use with the --model option.

The format for the model name is provider:model-name. For example:

  • openai:gpt-4o
  • anthropic:claude-3-5-sonnet-latest
  • google-vertex:gemini-2.0-flash

For a full list of supported models, see the PydanticAI documentation.

Default Values🔗

If not specified, Brag AI uses the following default values:

  • User: The owner of the GitHub API token, if provided
  • From date: None (no lower bound)
  • To date: None (the current date)
  • Limit: None (no limit)
  • Output: stdout (print to console)
  • Model: google-gla:gemini-2.0-flash
  • Language: English

Example Configurations🔗

Generate a brag document for contributions in 2023 using GPT-4o🔗

export OPENAI_API_KEY=your-openai-api-key
brag from-repo owner/repo --user github-username --from 2023-01-01 --to 2023-12-31 --model openai:gpt-4o

Generate a brag document in Portuguese using Claude🔗

export ANTHROPIC_API_KEY=your-anthropic-api-key
brag from-repo owner/repo --user github-username --language Português --model anthropic:claude-3-5-sonnet-latest

Save the brag document to a file and limit to the last 50 commits🔗

brag from-repo owner/repo --user github-username --limit 50 --output brag.md