Appearance
Moderation & abuse
Every Embed surface uses Zeeker Social's platform-wide and community moderation. Element presets do not create separate queues, roles, report types, support states, merchant actions, or review policies.
Layered model
┌─────────────────────────────────────────────────────────────┐
│ Platform-level (Zeeker trust & safety) │
│ • automated abuse detection │
│ • doxxing, CSAM, illegal content removal │
│ • global blocks (domains, IPs, content patterns) │
└─────────────────────────────────────────────────────────────┘
▲
│ feeds into
│
┌─────────────────────────────────────────────────────────────┐
│ Community-level (your admins + moderators) │
│ • rule citations on removals │
│ • per-community member bans │
│ • lock / unlock threads │
│ • slow mode │
└─────────────────────────────────────────────────────────────┘
▲
│ feeds into
│
┌─────────────────────────────────────────────────────────────┐
│ Embed-level (narrows the community capability) │
│ • require an allowed identity before participation │
│ • hide/disable allowed actions for this surface │
│ • apply per-user reply rate limits │
│ • allowed-origin whitelist │
└─────────────────────────────────────────────────────────────┘Per-widget knobs
In the configurator → Behavior and Advanced sections:
- Require sign-in — anonymous reads are still allowed, but reacting and replying require a Zeeker (or SSO) account.
- Disable replies — read-only widget; only reactions and shares.
- Disable reactions — useful for highly-moderated B2B contexts.
- Per-user reply rate limit — 1–120 replies / hour, default 30.
- Allowed origins — comma-separated list of hostnames. The widget refuses to render on others.
An embed can be stricter than its community but cannot re-enable an action, post type, chat destination, or content scope that the community disables. Community visibility and membership checks continue to run before embed policy.
Per-community moderation
Inside the admin center → Moderation tab, your moderators see:
- Open reports — reports queue from both
zeeker.comand the widget. - Removed content — log of what's been removed, by whom, with rule citations.
- Banned members — temporary or permanent bans, with appeals.
- Slow mode — minimum seconds between posts per user; toggle community-wide or per-thread.
- Locked threads — threads where replies are disabled (but reactions and saves still work).
Removals from the widget surface and from zeeker.com/c/<slug> are mirrored — moderators don't have to act in two places.
Question, review, poll, support-oriented, and ordinary feed presets all submit to this same moderation model because they use the same canonical Social posts and replies.
Reporting
Visitors can report a post or reply from the widget's overflow menu:
- Pick a reason (Harassment, Spam, Off-topic, Misinformation, Illegal, Other).
- Optionally add a note (≤ 280 chars).
- The report goes to your community's moderation queue + the platform queue.
Reports are anonymous to the reported author; reporters see a confirmation.
Spam mitigation
- Per-IP rate limits at the edge for unauthenticated requests.
- Posting cooldowns for new accounts (under 24h old): 5 minute minimum between posts in their first day.
- Domain blocklist at the link-share layer.
- Automated toxicity scoring on every new reply; high-confidence hits go straight to the mod queue rather than the public feed.
Platform actions
If a post breaks Zeeker's community standards, the platform team can act over the community's head:
- Hide a post globally (community admins see "Hidden by Zeeker T&S" in the modlog).
- Suspend a user account.
- Block a domain from being link-shared via any widget.
These actions are rare and always logged. Community admins are notified by email when one affects their community.
Audit log
Every moderation action — community or platform — is recorded in the modlog, accessible from Admin center → Moderation → Modlog. Filterable by actor, target, action type, and date.