Terms

Terms of service

Last updated: 2026-07-03

The short version: use MARVIN to moderate your server, don't use it to attack other people or the service itself, and remember that automated moderation is configured by your guild's admins — not by us. The long version follows.

1. Who these terms cover

“MARVIN” means the Discord bot and this web dashboard, operated by the person or team running this instance (the “operator”). These terms apply to anyone who interacts with the service: guild owners and admins who add the bot, moderators who use its commands and dashboard, and members of servers where the bot runs. Adding the bot to a server, signing into the dashboard, or using its commands means you accept these terms.

2. Discord's rules still apply

MARVIN runs on Discord, so Discord's Terms of Service and Community Guidelines apply on top of these terms. You must meet Discord's minimum age requirement (13, or higher where your country's digital-consent age says so) to use the service.

3. What the service is

MARVIN is a moderation tool: anti-spam, anti-scam-image detection, tickets, filters, role automation, and a dashboard to configure it all per guild. Detection is probabilistic — confidence scores, thresholds, perceptual hashes. That means false positives happen. The dashboard ships review queues so staff can confirm or reverse automated calls; the bot doesn't pretend to be a judge, it's a first responder.

4. Guild admins configure the moderation

Which filters run, how aggressive the thresholds are, and what happens on a match (warn, mute, kick, ban) is decided by each guild's admins in the dashboard. The operator provides the tool; your guild's staff decide how it's pointed. If you were actioned by the bot in a server, appeal to that server's staff — they hold the controls and the review queue.

5. Acceptable use

Don't:

  • use MARVIN to harass, target, or mass-action people — moderation tooling is for protecting communities, not raiding them;
  • probe, overload, or attack the service — no flooding the bot with junk to burn its rate limits, no attempting to bypass or poison its detection (e.g. deliberately seeding the scam-image denylist with innocent images);
  • attempt to access other guilds' data, other users' dashboard sessions, or operator-only surfaces;
  • resell or misrepresent the service as your own without the operator's consent.

Found a security hole? Report it privately via GitHub or the support server rather than exploiting it. Responsible disclosure is welcome; exploitation is a ban.

6. Your data

What the bot stores, for how long, and who can see it is covered in the privacy policy. Summary: moderation records live in the operator's own database, nothing is shipped to third-party analytics, and kicking the bot starts a 30-day purge window for your guild's data.

7. Paid plans

Where paid tiers are offered, they're billed per server through Stripe. Prices and what each tier unlocks are on the pricing page. Subscriptions renew until cancelled; cancelling stops future charges and the plan runs out at the end of the paid period. Downgrading never holds your data hostage — paid features switch off, your configuration and history stay.

8. Availability

The service is provided as-is, without uptime guarantees. The bot can go offline for deploys, host maintenance, or Discord outages — the status page tells the truth about it. Don't build anything safety-critical on the assumption MARVIN is always watching.

9. Termination

You can stop using the service at any time by kicking the bot — data deletion then follows the privacy policy's retention window. The operator can remove the bot from a guild, revoke dashboard access, or refuse service to anyone violating these terms or using the service to break Discord's rules.

10. Liability

To the maximum extent the law allows, the operator isn't liable for indirect or consequential damages from using the service — including moderation decisions made by the bot under your admins' configuration, missed detections, false positives, or data loss. If you paid for a plan, total liability is capped at what you paid in the last 12 months.

11. Changes

These terms get updated when the service changes. Each update bumps the “Last updated” date above; material changes get a changelog entry. Continuing to use the service after a change means you accept the new terms.

12. Contact

Questions about these terms go to the operator — via the support server linked in the footer (when configured) or GitHub issues.