Verification is only the first checkpoint
A member who completes a CAPTCHA can still spam, phish or abuse a community. New Member Shield protects the period immediately after verification while the member establishes a clean activity history.
Available now: join-burst observation
GuildWard can count deduplicated human joins inside a configurable guild-scoped time window. When the published threshold is reached, one aggregated incident appears in Activity and Audit and closes durably after the configured quiet period. This stage is observation-only: it does not kick, ban, restrict members or change channels automatically.
Joined → Verified → Probation → Clean activity → Trusted
↘ Reverification
↘ Manual ReviewConfigurable probation
Choose which roles, channels and actions are available during probation. Graduation can follow clean activity, elapsed time or another completed step. Trusted users, trusted roles and known bots can follow explicit exemption rules.
Probation and behavior-based enforcement belong to a further protection stage; they are not part of the current join-burst observer.
Deterministic protection
- Message-rate and duplicate-message controls
- Mention, link and invite limits
- Attachment-rate limits
- Discord account-age context
- Join-burst awareness
- Channel-spread detection
- Trusted-principal exemptions
These signals are understandable, testable and useful even when AI is unavailable.
Risk-Based Reverification
When configured behavior crosses a policy threshold, GuildWard can temporarily restrict access and request a proportionate step: another CAPTCHA, rules reconfirmation, linked-account check or Manual Review. The response stays connected to the policy and evidence that triggered it.
Pro Rejoin Guard
An explicit adverse or unresolved decision should not disappear because a member leaves before enforcement or review completes. Pro Rejoin Guard carries that decision into a new membership generation and can observe the return, keep the member Restricted, reopen review or deny a fresh journey.
Entries are server-scoped, time-bounded, auditable and appealable. They are not shared across GuildWard customers, and a rejoin alone is not treated as proof of malicious intent.
Reversible by design
GuildWard prefers reversible restrictions and review over irreversible automated punishment. Communities decide when clean activity restores access, when a moderator must review and how a member reaches support.
Smart Shield as an additional signal
Smart Shield can classify eligible probation activity for contextual scam patterns. AI informs policy; it does not replace deterministic controls, Discord permissions or human accountability.