feat: decouple read FeedId from inbound MailboxId

Separate the two feed identities so the public read URL never reveals the
inbound address and vice-versa:

- FeedId becomes an opaque high-entropy token (read id + KV key); MailboxId
  (noun.noun.NN) owns the inbound address and the untrusted-input boundary
  via MailboxId.parse. They map only through the inbound:<mailbox> secondary
  index, resolved solely at reception.
- inbound index lifecycle is owned by FeedRepository: written by save/saveConfig,
  dropped by removeFromList(Bulk) — symmetric, never mirrored by hand (removes the
  manual delete in feed-service + the cron loop, and a silent empty-catch).
- Feed.mailboxId exposes a MailboxId VO (symmetry with Feed.id); the
  mailbox@domain shape lives on MailboxId.emailAddress(domain).
- Distinguish mailbox_unknown (no feed claims the address) from feed_not_found
  (dangling index) for observability; both forwardable, both 404.
- Drop the redundant EmailParser.extractMailbox pass-through so MailboxId.parse
  is the single parse boundary.

Docs (README/INSTALL/CLAUDE.md/landing) and tests updated; 439 tests green,
tsc clean, build dry-run OK.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
This commit is contained in:
Julien Herr
2026-05-24 22:46:37 +02:00
parent f7f10779bc
commit 1a4a479190
43 changed files with 649 additions and 149 deletions
+9
View File
@@ -137,6 +137,15 @@ What gets forwarded vs dropped:
Expired feeds and blocked senders are dropped on purpose, so a real newsletter never leaks into your fallback inbox. Leave the variable unset to keep the original drop-and-log behavior.
### Inbound address vs feed URL
Each feed has **two independent identifiers**, on purpose:
- a friendly **inbound address** you subscribe newsletters with — `noun.noun.NN@yourdomain.com` (e.g. `apple.mountain.42@yourdomain.com`);
- an **opaque feed URL** for your reader — `https://yourdomain.com/rss/<random-id>` (also `/atom/<id>`, `/json/<id>`).
They are not derivable from each other. This means you can hand someone a feed URL without revealing the address that feeds it, and an address harvested by a newsletter sender can't be turned into your feed (requesting `/rss/<your-address>` returns 404). The admin dashboard shows both per feed; copy the address into signup forms and the feed URL into your reader. (Internally the inbound address is mapped to the feed by an `inbound:<address>` KV entry, resolved only when mail arrives.)
### Feed size limit
By default the worker keeps emails until the feed's stored data exceeds **512 KB**, then drops the oldest entries (and their KV records) to stay under the limit. This is more robust than a fixed entry count for HTML-heavy newsletters.