Monitor

Operations overview

Delivery, queue and reputation at a glance · auto-refreshes on the poll cycle

24-hour activity · 15-min buckets

● accepted ● delivered ● bounced ● deferred

Queue & connections

IP health (24 h)

Egress address	Identities	Delivered	Deferred	Bounced	Success

Sending identities (24 h)

EHLO / subdomain	IP	Deliv	Defer	Bounce	Success

Top recipient domains (24 h)

Domain	Deliv	Defer	Bounce

Bounce classification (24 h)

Recent failures (newest first)

no failures recorded in the last 24h

ISP throttling

System

Send

Identities

Sending domains, DNS publication and the panel host

Automatic DNS publication · Cloudflare

paste an API token once — every domain you add here gets its records (A, SPF, DKIM, DMARC, MX, TLS-RPT) created automatically. The panel only ADDS: existing records of your other websites stay untouched, and the panel can later remove only the records it created itself.

auto-publish

Panel host · HTTPS address

reach the panel with a real address instead of the raw IP: https://panel.yourdomain — the DNS record is created in Cloudflare, a genuine Let's Encrypt certificate is issued with your API token (DNS-01, no port 80 needed, never self-signed), and nginx serves it with automatic renewal. One domain → set up automatically. Several domains → choose here, and you can change or remove it any time.

subdomain:

Sending domains

every row is a sending domain with its own addresses, subdomain block and DMARC policy — add as many as you like (max 50 identities in total).

Domain	Identities	Addresses	DMARC

Advanced — preset server settings (open once if you need a relay)

outbound route (relay / direct)

queue behaviour (applies with APPLY)

DNS records to publish

Name	Type	Value	Why

Live DNS status

Identity	Address	A/AAAA	SPF	DKIM	DMARC	MX

checked live through 1.1.1.1 (TCP, DoH fallback — many providers block plain DNS) · fresh record changes can take up to the record TTL (~1 h) before they show here · MISSING = publish from the sheet on the left

Managed policy

Component	Status	Size

server-managed configuration — rewritten by APPLY, backed up automatically

Generated sheet (registrar copy)

load the page to generate the sheet…

Send

Campaigns

Performance analytics per campaign · opens, clicks and unsubscribes

Campaign performance

all idle sending completed archived

	Campaign	State	Sent	Delivered	Del%	Bounced	Bn%	Opens	Open%	Clicks	Click%	Unsub	Unsub%	Funnel

● delivered ● deferred ● bounced click a row for the per-recipient log

Recipient detail — select a campaign above

time	email	status	identity	response

spawns a fresh draft · "{name} · retry" · addressed to the failed + deferred log entries (newest first, cap 2000) — it starts idle, you press Start

Follow-up: max age (days)

the derived audience is a snapshot (audience_override) — it does not auto-update · suppressed and unsubscribed addresses are always excluded · the draft starts idle, you press Start.

archiving is for completed / stopped / failed campaigns — archived ones leave the default lists but stay selectable for follow-ups and compare

Engagement — all campaigns

opens / clicks are attributed to a campaign through the tracking id of every message · unsubscribes are RFC 8058 one-click and auto-suppress the address.

Send

Composer

Draft, audience and delivery run

Composer

click a token to insert it at the cursor — substituted per recipient at send time · subscriber tags render comma-separated

a plain-text part improves delivery

Templates

templates store subject + body + text part · Load fills the form · names must be unique

Audience

subscribers: – · mailing now: – · suppressed: –
adds/removes happen in the SUBSCRIBERS tab

Delivery controls

messages/minute

track opens track clicks unsubscribe footer

audience lists

Main (default — every subscriber)

delivery window –

off — sends around the clock · server time (UTC)

starts automatically at this time · server time (UTC) — safe mode holds it

off (single subject)

0 drafts

Preflight · campaign readiness

not run

informational only — failing checks don't block save; Start asks once per draft

Run

IDLE

Live results (newest first — last 200)

time	email	status	identity	SMTP response

Send

Audience

Segments, subscribers, imports and suppression

Segments

Audience lists

Import subscribers (paste)

the scraper writes domain,email rows — upload the file here or point the console at the path on this server (readable folders only). dropping a file on the upload button works too.

Maintenance

tip: bounces and unsubscribes are suppressed automatically

Import history

when	source	list	added	dupes	invalid	by	action

undo removes the members a batch added — anyone already in an earlier batch stays

Audience health

not checked yet

Subscribers

	#	email	name	tags	lists	engagement	added

page 1 of 1

Suppression list

#	email	kind	reason	when	action

Monitor

Deliverability

DNS, blocklists and destination health · inbox readiness

Identity DNS checks

Identity	Address	A/AAAA	SPF	DKIM	DMARC	MX

checks run against the authoritative nameservers — results cached 5 min · also verify with sudo ./setup-multi.sh --check-dns

IP reputation & blocklists

Egress address	Spamhaus ZEN	Barracuda	SpamBarrel	SORBS	Verdict

Spamhaus refuses queries from public resolvers (1.1.1.1 / 8.8.8.8) and answers "query blocked" instead of a verdict - check it from your own resolver or on spamhaus.org. A listing here is a reputation blocker: fix it before sending.

Destination health (who is refusing us)

Domain	Deliv	Defer	Bnc	Bnc%	Verdict

Failure categories (24 h)

critical = bounce rate ≥ 10% · fix that domain first

Latest refusal samples

no failures recorded in the last 24h

Warm-up playbook (2026)

day 1-2    : ~150-300 msgs / address      (spread over the day)
day 3-5    : ~600-1200 msgs / address
week 2     : double every 2 days to steady state
watch      : IP health table — pause any address with rising 5xx
rules      : real subject lines · working unsubscribe · no misleading headers
             keep list hygiene (this panel auto-suppresses bounces)

Monitor

Alerts

Rules, channels and notification history

Active alerts

Rules

On	Alert	Metric	Threshold	Severity	Current

the engine re-evaluates every 60 seconds and fires a notification when a rule changes state (ok → firing → ok).

Notification channels · webhook & e-mail

the e-mail is delivered through your own private relay (authenticated SMTP) — no third party involved.

Telegram

silent (no sound)

alerts campaigns daily digest

1) open @BotFather → /newbot → copy the token · 2) send your bot a message (or add it to the channel) · 3) get your chat id from @userinfobot → paste it above. The token stays on the server — the browser only ever sees the last four characters.

Alert history

time	alert	severity	value	state	notified

Operate

Operations

Diagnostics, safe mode and message queues

Diagnostics — nothing is sent unless you press the button

injects one message through the local SMTP listener with DKIM signing and queueing — it is never sent automatically, not at boot, not on a schedule.

bounces leftover test messages (probe/mail-tester) that are stuck in the scheduled queue because port 25 is blocked. Only test queues are touched.

Production readiness

State storage

—

Safe mode

OFF

Safe mode stops every campaign immediately — the cursor, the counts and the queue stay intact, so switching it off resumes exactly where it stopped.

Warm-up mode

OFF

While warm-up mode is on, delivery and connection rates are halved for the first 14 days of a new address or server. Destination profile behaviour stays identical — switching it off restores full rates.

Message queues

Queue	Messages

PURGE bounces every message of that queue immediately (AdminBounce) and stops new arrivals for 5 seconds.

Administrative rules

loading…

Panel doctor — self-diagnostics

not run yet — checks storage, services, certificate, queue and reachability

repairs:

every repair asks for confirmation and lands in the audit trail. Resync re-imports only legacy files that are newer than the stored documents (degraded-window recovery).

Backup & restore

backups carry every stored document, the audit trail and a redacted config copy.

restore replaces the documents present in the zip and leaves everything else alone; the console config file is never restored (config stays managed by Settings) and the audit trail is append-only. In postgres mode also run pg_dump for a full binary backup.

Console preferences

auto refresh

preferences live in this browser only · the console remembers the last view you opened.

Telegram

not configured

alerts that fire, campaigns that start or finish and the optional daily digest are pushed to your chat. Configure it under Alerts.

Managed configuration

select an item…

policy and console configuration live server-side — this list shows placement and size only

Operate

Settings

Console configuration, self-update and restarts — admin only

Console configuration

Public base URL

Auto refresh (seconds, 5–600)

Daily send cap (0 = off)

engine-wide sends per UTC day; when the cap is reached, sending campaigns auto-pause with a CAP badge until the next day or until the cap is raised. Test sends never count.

Probe domains (comma separated)

tracking — — read-only here; it follows the sending domain settings in Identities.

SMTP host (local injection)

SMTP port

send_test flag

Database dsn (stored value is masked)

leave masked/empty to keep the stored dsn · a test connect never saves anything · a saved dsn change needs the restart (a bad dsn falls back to file mode, the panel stays up)

Saving rewrites the console configuration — a timestamped backup copy is kept first — and restarts the console. This page reconnects automatically when the panel is back. Telegram has its own card under Operations (configure in Alerts).

Console update

loading…

Console core - validated with a python syntax check before anything is touched

Console interface - sanity-checked: doctype, sign-in element, closing tag

the bash installer is only needed for the first deployment — updates, configuration and maintenance live here. A failed validation never touches the running files, and every accepted swap keeps a rolling backup under state/backups/ (last 5 kept).

Danger zone

a console restart takes ~2 seconds and this page reconnects on its own; an engine restart bounces in-flight deliveries back to the spool recovery. Both actions are audited.

Operate

Team & audit

Operators, roles and the audit trail

My account · change password

use at least 10 characters — a mix of upper/lowercase, digits and symbols reads “strong”.

changing the password keeps this session signed in and signs your other sessions out. Every role can change their own password here; the team table below is admin-only.

Team & access

User	Role	Status	Created

Add operator

every mutating action is written to the audit trail below.

Audit trail

time	actor	action	target	detail

Monitor

Events

Sends and engagement tracking

–

opens (total)

–

clicks (total)

–

unsubscribes (total)

Recent events (sends / tracking)

time	kind	detail	response

Raw delivery-log failures (24 h)

none

Operations overview

24-hour activity · 15-min buckets

Queue & connections

IP health (24 h)

Sending identities (24 h)

Top recipient domains (24 h)

Bounce classification (24 h)

Recent failures (newest first)

ISP throttling

System

Identities

Automatic DNS publication · Cloudflare

Panel host · HTTPS address

Sending domains

Add domain

DNS records to publish

Live DNS status

Managed policy

Generated sheet (registrar copy)

Campaigns

Campaign performance

Recipient detail — select a campaign above

Compare campaigns

Engagement — all campaigns

Composer

Composer

Templates

Draft history

Audience

Delivery controls

Preflight · campaign readiness

Run

Live results (newest first — last 200)

Audience

Segments

Audience lists

Import subscribers (paste)

Maintenance

Import history

Audience health

Subscribers

Subscriber

Suppression list

Deliverability

Identity DNS checks

IP reputation & blocklists

Destination health (who is refusing us)

Failure categories (24 h)

Latest refusal samples

Warm-up playbook (2026)

Alerts

Active alerts

Rules

Notification channels · webhook & e-mail

Telegram

Alert history

Operations

Diagnostics — nothing is sent unless you press the button

Production readiness

State storage

Safe mode

Warm-up mode

Message queues

Administrative rules

Panel doctor — self-diagnostics

Backup & restore

Console preferences

Telegram

Managed configuration

Settings

Console configuration

Console update

Danger zone

Team & audit

My account · change password

Team & access

Add operator

Audit trail

Events

Recent events (sends / tracking)