Every poll's outcome has always been written to source_health and shown on
/status, but only to someone who went and looked. Nobody did: Siliconera has
been refused on 3,769 polls in a row and Digital Foundry on 1,900, and
neither has ever fetched successfully. SendAdminWarning existed and nothing
called it.
A watcher now reads source_health every 15 minutes and sends one notice to
matrix.admin_room covering everything that changed since it last spoke:
feeds that have failed 5 polls in a row, feeds that have fetched since they
were reported down, and feeds that fetch fine but have kept nothing new for 7
days (every healthy source produces something inside about 4). What has been
said is stored on the row (down_alerted_at, quiet_alerted_at), so each
episode is reported once and a restart doesn't repeat it. A notice that
fails to send is not marked, and goes out on the next pass.
last_new_at is new, moved only when a poll keeps a story. Existing rows start
from their newest retained story, or from now if they have none, and a new
source starts at its first poll: nothing on disk says whether a source with
no stories is long dead or was added yesterday, so it gets a full window
before it can be called quiet.
With no admin room configured the watcher doesn't run at all, rather than
marking alerts as delivered to nowhere.
The source tag credits an outlet Pete is relaying (`ars technica`). On
his own reporting it rendered as a trailing `pete` under a message he
already sent - him signing his own name. Drop Source on adventure posts
and omit the tag line entirely when there's nothing to credit; RSS posts
keep theirs.
- Route image-validation, Matrix image-download, and wayback/archive.today
fetches through safehttp so feed-controlled URLs can't reach loopback/
RFC1918/cloud-metadata IPs (incl. via redirects).
- Cap feed body size with safehttp.LimitedBody (16 MiB) to prevent OOM.
- Fail closed if matrix.pickle_key is unset/<16 chars; drop the hardcoded
"pete_pickle_key" default that silently weakened E2EE-at-rest.
- Gate !post behind a matrix.admins allowlist; empty = disabled (channel
rooms are public, so empty must not mean anyone).
- Reject /\host open-redirect bypass in post-login safeNext.
- Allowlist http(s) schemes for the Matrix link href; escape JSON embedded
in inline <script> (prefs/sources blobs).
Two bugs prevented Pete's device from ever cross-signing itself:
1. Double login / split-brain: New() did a manual mx.Login() AND set
ch.LoginAs, so cryptohelper.Init() logged in a second time on a fresh
crypto store, minting a separate device. device.json recorded one device
while the olm account belonged to another, and every cold start leaked an
orphan device. Pete's outgoing events were attributed to a device whose
keys no client could verify. Removed ch.LoginAs (auth is already handled by
the manual login + isTokenValid re-login path).
2. Unreachable bootstrap: the reset gate used
IsDeviceTrusted(mach.OwnIdentity()), but OwnIdentity() hard-codes
Trust=TrustStateVerified, so it always returns true and the bootstrap/reset
branch was never entered. Replaced with GetOwnVerificationStatus(), which
actually checks whether the current device key is signed by our
self-signing key.
- Detect explicit paywall markers (article:content_tier meta, JSON-LD
isAccessibleForFree) so metered articles fall back to Wayback even
when body length is above threshold
- Blank-import go.mau.fi/util/dbutil/litestream so the sqlite3-fk-wal
driver mautrix's cryptohelper depends on is registered
- Fix /img route: Go ServeMux requires {wildcard} to be a whole segment,
so capture {name} and strip the .avif suffix in the handler
- !post falls back to newest unposted story for the channel when the
in-memory queue is empty (the steady state under round-robin).
- Accept ❓️/❔️ (U+FE0F variation selector) as question reactions —
the bare codepoints alone missed clients that render the colored emoji.
- Rewrite Guardian i.guim.co.uk thumbnails to width=1200 so we stop
rejecting real images as "tracking pixels"; relabel the size warning.
- Log decrypt failures and reactions on events not in post_log so future
silent drops surface instead of vanishing.
Typed in a configured channel room, !post pops the head of that channel's
queue and sends immediately, bypassing min-interval, burst cap, and daily
cap. Canonical-URL dedup still applies. Empty queue gets a threaded reply.
Four related fixes after Pete flooded a channel and ignored Ctrl-C:
1. Global daily cap (posting.daily_cap_total, default 5): hard ceiling on
posts across ALL channels in a rolling 24h window. Checked before the
per-channel min-interval and burst-cap.
2. Shutdown no longer flushes the queue. Previous drainAll posted every
remaining item with rate limits disabled — which was literally the
flood. Replaced with dropOnShutdown that clears queues and logs
the count.
3. Poller respects ctx mid-loop. pollOnceWithErr now takes ctx and bails
between items, so Ctrl-C doesn't have to wait for ~30s of network
per pending story before shutdown can complete.
4. Double-image fix. PostStory now reports imageSent; the queue clears
ImageURL before retry so a text-send failure after a successful
image upload doesn't re-post the image.
When a user reacts ❓ on one of Pete's posts, fetch the article body,
ask Ollama for a 3-bullet summary, and post it as a threaded reply
rooted at the original story event. Per-process cooldown of 5min per
story keeps repeated reactions from re-summarizing.
- ingestion.FetchArticleBody: visible <p> text capped at 8000 chars
- classifier.OllamaClient.GenerateText: non-JSON variant
- storage.GetStoryByGUID: full row lookup
- matrix.PostThreadedReply: m.thread + m.in_reply_to fallback
- poster.SetReactionCallback: optional hook fired after recording
- New package: internal/explainer
prosolis