Skip to content

Step 3: Configure APIs

Configure API access for GitHub and your AI provider.

Time estimate: 10-15 minutes

Complete these tasks

  • Create and test GitHub API token
  • Obtain and configure AI provider API key
  • Review security best practices
  • Verify API connections

Get API keys

1. GitHub token (required)

Fetch commits and pull requests from repositories with this token.

Free for public repos, included with GitHub account for private repos.

Create a GitHub token

  1. Navigate to github.com/settings/tokens

  2. Click Generate new token and choose Generate new token (classic)

  3. Configure the token:

  4. Note: "Release Notes Automation" (describes purpose)

  5. Expiration: Choose based on your security policy (90 days recommended)

  6. Scopes: Select the minimum needed:

    • For public repos only: public_repo
    • For private repos: repo (full control)

  7. Click Generate token and copy it immediately. The token format is ghp_xxxxxxxxxxxxxxxxxxxx.

Store the token in your password manager immediately. You cannot view it again after leaving the page.

2. AI provider API key (required)

Choose one provider:

Claude excels at following detailed instructions and explaining reasoning.

Get an Anthropic API key:

  1. Go to console.anthropic.com
  2. Sign up or log in
  3. Click Get API Keys in the navigation
  4. Click Create Key
  5. Name it "Release Notes Automation"
  6. Copy the key (starts with sk-ant-)

Pay-as-you-go pricing: approximately $0.01-0.03 per release notes generation

Free tier: $5 credit for new accounts

Option B: OpenAI (GPT)

GPT offers wide availability, fast performance, and good general capabilities.

Get an OpenAI API key:

  1. Go to platform.openai.com
  2. Sign up or log in
  3. Click your profile then View API Keys
  4. Click Create new secret key
  5. Name it "Release Notes Automation"
  6. Copy the key (starts with sk-)

Pay-as-you-go pricing: approximately $0.02-0.05 per release notes generation

Free tier: $5 credit for new accounts (first 3 months)

Configure your keys

1. Open configuration file

Edit config.yaml in the repository root:

# Navigate to repository root
cd /path/to/docs-automation-examples

Open in your text editor:

# VS Code
code config.yaml

# Cursor
cursor config.yaml

# Nano (terminal)
nano config.yaml

2. Add your keys

For Anthropic (Claude):

# AI Provider
ai_provider: "anthropic"
ai_api_key: "sk-ant-your-actual-key-here"
model: "claude-3-sonnet-20240229"

# GitHub
github_token: "ghp_your-actual-token-here"

# Optional: Set a default repository
default_repo: "your-username/your-repo"
output_file: "release_notes.md"

For OpenAI (GPT):

# AI Provider
ai_provider: "openai"
ai_api_key: "sk-your-actual-key-here"
model: "gpt-4"

# GitHub
github_token: "ghp_your-actual-token-here"

# Optional: Set a default repository
default_repo: "your-username/your-repo"
output_file: "release_notes.md"

3. Save the file

Save and close your editor.

The config.yaml file is in .gitignore to prevent accidental commits. Never commit this file. Verify it does not appear in git status:

git status
# config.yaml should not appear as modified

Test your configuration

Test 1: Configuration file loads

cd 01-release-notes-automation
python -c "import yaml; config = yaml.safe_load(open('../config.yaml')); print('Configuration loaded')"

The output shows Configuration loaded.

Test 2: GitHub API access

python -c "
from github import Github
import yaml
config = yaml.safe_load(open('../config.yaml'))
g = Github(config['github_token'])
user = g.get_user()
print(f'GitHub API working. Connected as: {user.login}')
"

The output shows GitHub API working. Connected as: your-username.

Test 3: AI API access

Test Anthropic:

python -c "
import anthropic
import yaml
config = yaml.safe_load(open('../config.yaml'))
client = anthropic.Anthropic(api_key=config['ai_api_key'])
message = client.messages.create(
    model='claude-3-sonnet-20240229',
    max_tokens=10,
    messages=[{'role': 'user', 'content': 'Say hello'}]
)
print('Anthropic API working')
print(f'Response: {message.content[0].text}')
"

Test OpenAI:

python -c "
from openai import OpenAI
import yaml
config = yaml.safe_load(open('../config.yaml'))
client = OpenAI(api_key=config['ai_api_key'])
response = client.chat.completions.create(
    model='gpt-4',
    messages=[{'role': 'user', 'content': 'Say hello'}],
    max_tokens=10
)
print('OpenAI API working')
print(f'Response: {response.choices[0].message.content}')
"

The output shows a success message with a response.

Follow security best practices

Do these actions

  • Use environment variables for production: 

    export GITHUB_TOKEN="your-token"
export AI_API_KEY="your-key"

  • Add config.yaml to .gitignore (already done)

  • Rotate keys regularly (every 90 days)

  • Use minimal scopes (only public_repo when possible)

  • Store in password manager for backup

Avoid these actions

  • Never commit config.yaml with real keys
  • Share API keys in Slack or email
  • Use production tokens for testing
  • Hardcode keys in scripts
  • Push to public repositories with keys

Troubleshooting

Invalid GitHub token

Error: 401 Unauthorized or Bad credentials

Solutions:

  1. Verify token is copied correctly (no extra spaces)
  2. Check token has not expired
  3. Verify required scopes are enabled
  4. Try regenerating the token

Invalid AI API key

Error: Invalid API key or Authentication failed

Solutions:

  1. Verify key is copied correctly
  2. Check you are using the right provider (anthropic versus openai)
  3. Verify billing is set up (after free tier)
  4. Check API key is active in provider console

Configuration file not found

Error: FileNotFoundError: config.yaml

Solutions:

  1. Verify you are in the right directory
  2. Check the file exists: ls -la config.yaml
  3. Verify you copied from config.example.yaml

Module import errors

Error: ModuleNotFoundError: No module named 'anthropic'

Solutions:

  1. Activate virtual environment: source venv/bin/activate
  2. Reinstall dependencies: pip install -r requirements.txt
  3. Verify Python version: python --version (3.8+)

Summary

You created a GitHub API token with appropriate scopes, obtained an AI provider API key, configured both keys securely, verified API connections work, and reviewed security best practices.

Next step

Run your first automation with APIs configured.

Next: Step 4 - First run

Estimated cost per run: GitHub API is free (rate limited to 5,000 requests per hour), AI API costs $0.01-0.05 per release notes generation. For typical biweekly releases: approximately $0.50-1.00 per month.