◆STATUS · SCANNERS · APIS · CRONS · UPTIME
Is the watchdog healthy?
Every receipt on this site is generated by a small cron of scanners running on Vercel + GitHub Actions. This page is the public read-out of that infrastructure - what's online, what's queued, when it last ran, what broke in the last 30 days, and which optional integrations are off (graceful no-op vs. actual outage). Auto-refreshes every 60s.
100
CORE SYSTEMS · OPERATIONALAll systems normal10 of 10 user-facing services healthy. 0 optional integrations are unconfigured (graceful no-op) and don't affect site availability.
UPTIME · 30D100.00%across user-facing services
RESPONSE · P95—not measured at this tier
LAST INCIDENT—none in 30d
NEXT PRICING SCAN09:25:0214:00 UTC daily
last checked 0s ago·refresh auto · 60s·checks run inside Vercel·history retained 90 days
§ 01SERVICES10
OPERATIONAL· 10
Public site (gotnerfed.com)Next.js 15 + Vercel edge + ISR
p95 · 1h—
OPERATIONAL
Public API v1/api/v1/health · tier-aware rate limits
p95 · 1h—
OPERATIONAL
RSS feed/feed.xml · regenerates every 60s
p95 · 1h—
OPERATIONAL
Anthropic Haiku 4.5 (classifier)AI Jury classification active · falls back to heuristic if rate-limited
p95 · 1h—
OPERATIONAL
Resend (transactional email)magic-link sign-in + weekly digest delivery
p95 · 1h—
OPERATIONAL
Bluesky botauto-broadcasts critical/major receipts
p95 · 1h—
OPERATIONAL
Upstash Redis (KV cache)cron logs + sentiment snapshots + rate-limit counters
p95 · 1h—
OPERATIONAL
Postgres (Better-Auth + subscriptions)auth + Stripe sub state + watchlist + webhook registry
p95 · 1h—
OPERATIONAL
Stripe (paid subscriptions)paid-tier provisioning live · webhooks verified
p95 · 1h—
OPERATIONAL
Web push notificationsVAPID-signed delivery to subscribed clients
p95 · 1h—
OPERATIONAL
§ 02SCANNER CADENCE6
When the watchdog runsAll times UTC · current: 04:34:57 · next pricing scan in 9h 25m
NOW · 04:3414:00vendor pricing scanon-demandreddit sentiment scan
0003060912151821
14:00 UTCVendor pricing scan · daily across tracked vendorsin 9h 25m
on-demandReddit sentiment scan · Claude Haiku 4.5 · cached daily in KVon first vendor render
chainedWatchlist forecast alerts · fires after scan-vendors completesafter next scan
chainedGitHub alt-discovery · fires after scan-vendors on UTC Mondaysafter next Mon scan
on diffAI Jury classification · fires per detected diff via /api/jury/[id]event-driven
on receiptBluesky bot broadcast · fires on every critical/major receiptevent-driven
§ 03LIVE CRON HEALTH13
What actually ran in the last cycleEach cron writes a timestamp to KV on every successful run. Stale = missed its expected interval by 50%; down = missed by 200% or recorded a failure.
Vendor pricing scandaily 14:00 UTC · scan complete
14h agoHEALTHYReddit sentiment (on-demand)on-demand (inline cache) · 21 vendors snapshotted
88d agoHEALTHYWatchlist forecast alertsdaily 14:30 UTC · 0 sent · 1 cooldown · 2 no-trigger · 0 no-email
14h agoHEALTHYGitHub alt-discovery scanweekly Mon 09:00 UTC · scan complete
3d agoHEALTHYRefresh alt star countsad-hoc / weekly · stars refreshed
3d agoHEALTHYNerfBench leaderboarddaily (configurable) · 1 scored, 0 skipped · subset=fast · cursor rotated
14h agoHEALTHYLiteLLM price diffdaily 14:00 UTC (chained) · diff complete · 3002 litellm
14h agoHEALTHYVendor blog announcementsdaily 14:00 UTC (chained) · 22/24 feeds ok, 10 new entries, 0 persisted, 0 releases (0 tweeted), 0 broadcast (2 flaky: perplexity,midjourney)
14h agoHEALTHYWeekly forecast → Bluesky/Xweekly Sun 16:00 UTC · posted top 1: Gemini (Google)(93%)
4d agoHEALTHYFreeze weekly Receipt issueweekly Mon 14:00 UTC (chained) · issue vol-i-issue-031 frozen - 1 receipts + broadcast queued
3d agoHEALTHYSite health digest emailevery 2 days (chained) · 0 cron · 6 announce · emailed=false
14h agoHEALTHYCD proxy rotationdaily (chained) · 0 rotated / 0 errored / 33 watches
14h agoHEALTHYWeekly round-up postweekly Sun 16:00 UTC (chained) · draft · 20 events · 911w · gen=ok-5sec · emailed=true · tweeted=false
4d agoHEALTHY§ 04SUBSCRIBE TO STATUS UPDATES
Get notified when something breaks
We don't run a SaaS status-page provider - too easy to fake. Subscribe to one of these public channels instead:
@gotnerfed.bsky.socialweekly digest + incident posts~2-3/wk→GitHub issues trackerincident threads + post-mortemson incident→RSS firehoseevery receipt + status eventlive→
Status checks run inside Vercel - for true third-party uptime, see the linked external monitors in the repo README.
incidents · 30d0total
-30dnow