Prompt for creating a small CLI "read it later" application in Go

PROMPT.md

Task

We’re going to write a “read it later”-tool (ril) in Go, and test it using the BATS (Bash Automated Testing System). Its purpose is to save URLs to an SQLite database for later viewing.

Functionality

Adding URLs

Each URL is assigned an integer id. IDs are formatted right-aligned in fields of width 4. The ID used in the examples below is 17.

A newly added URL is retrieved and the <title> element is parsed (e.g. https://www.example.com -> Example Domain). Use a User-Agent string containing “Read-it-later”. If the URL does not resolve with a 200 status, use “[Unknown Title]” as the title.

New URL

$ ril https://www.example.com
* [  17] added. 
  Example Domain
  https://www.example.com

Multiple new URLs

$ ril https://www.example.com https://www.example.org
* [  17] added.
  Example Domain
  https://www.example.com
  
* [  18] added.
  Example Domain 2
  https://www.example.org

$ ril foo bar
* Error: `foo` is not a valid URL
* Error: `bar` is not a valid URL

Existing URL

$ ril https://www.example.com
* [  17] (previously added on 2025-12-19)
  Example Domain
  https://www.example.com

Existing URL (already read)

$ ril https://www.example.com
* [  17] (previously read on 2025-12-19, marked unread)
  Example Domain
  https://www.example.com

Existing URL (Snoozed)

$ ril https://www.example.com
* [  17] (exists, snoozed until 2026-01-11)
  Example Domain
  https://www.example.com

Marking URLs as read

$ ril 17
* [  17] Marked as read!
  Example Domain
  https://www.example.com

$ ril read https://www.example.com <alias: ril r>
* [  17] Marked as read!
  Example Domain
  https://www.example.com

Deleting URLs

$ ril rm 17 <aliases: ril remove, ril delete>
* [  17] REMOVED
  Example Domain
  https://www.example.com

Listing URLs

The bare command lists the five oldest, unread and un-snoozed URLs.

$ ril
* [  17]
  Example Domain
  https://www.example.com

* [  18]
  Example Domain
  https://www.example.org

* <lists up to 3 more>

$ ril -n 20
<lists the 20 oldest unread and unsnoozed URLs>

Opening a Random URL

$ ril open <aliases: ril lucky, ril o>
* [  17]
  Example Domain
  https://www.example.org

<web browser opens with that URL>

Snoozing URLs

Defaults to snoozing for 7 days.

$ ril snooze 17 --days 14 
* [  17] Snoozed until 2026-01-11
  Example Domain
  https://www.example.com

If the URL is already snoozed, the earlier of the snooze dates takes effect.

Undo

Adding

$ ril https://www.example.com
* [  17] 
  Example Domain
  https://www.example.com

$ ril undo
* [  17] REMOVED
  Example Domain
  https://www.example.com

Marking read

$ ril 17
* [  17] Marked as read!
  Example Domain
  https://www.example.com

$ ril undo
* [  17] Marked as unread!
  Example Domain
  https://www.example.com

Snoozing

$ ril s 17
* [  17] Snoozed until 2026-01-11
  Example Domain
  https://www.example.com

$ ril undo
* [  17] Marked as ready!
  Example Domain
  https://www.example.com

Note that it is fine to lose the previous snoozed_until date when undoing a snooze that modified an already-snoozed URL.

Database

The database is stored in ~/.local/share/ril/ril.db, but the location can be overridden using the --database/-D flag.