Chronobsky

Archives your Bluesky posts and replies locally, and generates a static blog to browse them chronologically. Uses no external dependencies, just python3.

1. fetch_bsky.py - Crawls and archives your Bluesky posts and replies into local JSON files.
2. gen.py - Generates a static website from the archived data with chronological navigation, dark/light theme, and responsive design.
3. deploy_vercel.py - Deploys the generated site to Vercel as a production deployment.

.env

BLUESKY_HANDLE=yourhandle.bsky.social
BLUESKY_APP_PASSWORD=xxxx-xxxx-xxxx-xxxx
VERCEL_TOKEN=your_vercel_api_token

• BLUESKY_HANDLE + BLUESKY_APP_PASSWORD -- app password from Bluesky Settings > App Passwords
• VERCEL_TOKEN -- API token from https://vercel.com/account/tokens (only needed for deploying)

Note

Be careful with VERCEL_TOKEN -- these tokens are account-level and don't have project-level granularity. Consider creating a separate Vercel account for this project so your other projects remain safe in case of a token compromise.

Usage

$ python3 fetch_bsky.py     # fetch new posts from Bluesky
$ python3 gen.py            # generate static site in dist/
$ python3 deploy_vercel.py  # deploy dist/ to Vercel
$ ./run.sh                  # run all three in sequence

Deployment

1. Clone the repository:

   git clone --depth=1 https://github.com/toxdes/chronobsky.git
   cd chronobsky
   

2. Get your Bluesky app password:

   • Go to Bluesky Settings > App Passwords
   • Click "Add App Password", give it a name, and copy the generated password
   • Your handle is your Bluesky username (e.g. user.bsky.social or user.com if you have a custom domain)

3. Get your Vercel API token:

   • Go to https://vercel.com/account/tokens
   • Create a new token with any name and copy it

4. Set up the environment:

   cp .env.example .env
   

   Then edit .env and fill in your handle, app password, and Vercel token.

5. Register a cron job to run every 48 hours:

   crontab -e
   

   Add this line:

   0 6 */2 * * /path/to/chronobsky/run.sh
   

   This runs the full pipeline (fetch, generate, deploy) at 6 AM every other day.