Cloudflare Worker that stores a curated reading list in KV and exposes JSON, Markdown, and RSS endpoints.

• Cloudflare account
• Node.js 18+
• Wrangler CLI

Install Wrangler:

npm install -g wrangler

Login:

wrangler login

Minimal HTML reading log page (use for your subdomain).

Optional query params:

• limit (default 5, max 20)

Returns JSON list of items (newest first).

Query params:

• limit (default 5, max 20)

Returns a Markdown list for README embedding.

Query params:

• limit (default 5, max 20)

Returns an RSS feed.

Query params:

• limit (default 5, max 20)

Adds an item. Requires auth.

Headers:

• Authorization: Bearer $READING_TOKEN or X-Reading-Token: $READING_TOKEN

JSON body:

{
  "title": "Article title",
  "url": "https://example.com/article",
  "added_at": "Optional ISO timestamp"
}

Removes an item by URL. Requires auth.

Headers:

• Authorization: Bearer $READING_TOKEN or X-Reading-Token: $READING_TOKEN

JSON body:

{
  "url": "https://example.com/article"
}

Returns 404 if no item matches the given URL.

• If title length is 85+, truncate to 80 characters and append ….
• If title length is 84 or less, display the full title.

1. Create a KV namespace and bind it in wrangler.toml as READING_KV.

   wrangler kv namespace create READING_KV
   wrangler kv namespace create READING_KV --preview

2. Add the KV IDs to wrangler.toml.

3. Optionally store your token locally in .env (do not commit it):

   cp .env.example .env
   # edit .env and set READING_TOKEN

4. Set the write token in Cloudflare:

   wrangler secret put READING_TOKEN

   Hint: Use a long random string, e.g. openssl rand -hex 32.

   You will be prompted to create a Cloudflare Worker if one does not exist yet. Select "yes" to create one.

5. Optional metadata (set in wrangler.toml under [vars] or via secrets):

   wrangler secret put READING_SITE_TITLE
   wrangler secret put READING_SITE_URL
   wrangler secret put READING_MORE_URL

6. Deploy:

   wrangler deploy

curl -X POST https://YOUR_WORKER_URL/reading/add \
  -H "Authorization: Bearer $READING_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"title":"Example title","url":"https://example.com"}'

curl -X DELETE https://YOUR_WORKER_URL/reading/remove \
  -H "Authorization: Bearer $READING_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"url":"https://example.com/article"}'

If your GitHub README contains:

<!-- READING-LOG:START -->
...
<!-- READING-LOG:END -->

You can populate it with the Markdown endpoint:

https://YOUR_WORKER_URL/reading/markdown?limit=5

Point your reading.<domain> to the Worker and it will serve the HTML page at /.

Recommended (Worker Route):

1. DNS record: create a CNAME record for reading that points to your root domain (@) or any existing hostname in your zone. Make sure it’s proxied (orange cloud).
2. Worker route: in Cloudflare → Workers & Pages → your worker → Triggers, add a route:

   reading.yourdomain.com/*
   

Alternative (Custom Domain): If your Cloudflare plan supports Worker custom domains, add reading.yourdomain.com in the worker’s Settings → Domains and Cloudflare will create the DNS record automatically.

Your RSS remains available at:

https://reading.yourdomain.com/reading/rss