Get started →
How it works Features Bot detection Integrations Pricing Insights FAQ Log in Get started free →

// AI Content Moderation for Communities

Stop the
noise.
Protect
your people.

Sentinel catches grooming, scams, bots, member poaching, and harmful content that generic tools completely miss — in under 100ms.

Start free — 50 checks → See how it works ↓

// No credit card required · 14-day trial on paid plans

92%
Detection accuracy
<2%
False positive rate
$2k+
Avg scam prevented
6 hrs
Saved per week
BLOCKED · "Send me $50 via CashApp, lost my wallet" — financial_scam · 98% WARNED · "DM me, I run a better program than this" — poaching · 91% APPROVED · "Great post! Really helpful breakdown of Module 3" — clean BLOCKED · "You seem mature for your age, want to chat privately?" — grooming · 97% BOT RISK · User3847291 — score 78/100 — 3 flags in 10 min · new account WARNED · "Join my Telegram, way better content there" — off_platform · 84% APPROVED · "Has anyone tried the homework from week 2?" — clean BLOCKED · "Send 0.1 ETH to get 0.2 back guaranteed" — crypto_scam · 99% CAMPAIGN · Same script hitting 4 communities — coordinated bot attack detected APPROVED · "Question about tomorrow's call — what timezone?" — clean BLOCKED · "Send me $50 via CashApp, lost my wallet" — financial_scam · 98% WARNED · "DM me, I run a better program than this" — poaching · 91% APPROVED · "Great post! Really helpful breakdown of Module 3" — clean BLOCKED · "You seem mature for your age, want to chat privately?" — grooming · 97% BOT RISK · User3847291 — score 78/100 — 3 flags in 10 min · new account WARNED · "Join my Telegram, way better content there" — off_platform · 84% APPROVED · "Has anyone tried the homework from week 2?" — clean BLOCKED · "Send 0.1 ETH to get 0.2 back guaranteed" — crypto_scam · 99% CAMPAIGN · Same script hitting 4 communities — coordinated bot attack detected APPROVED · "Question about tomorrow's call — what timezone?" — clean

// How it works

Three ways to connect.
One decision engine.

Connect your community in minutes. Every integration uses the same AI engine with the same 36+ detection rules and bot scoring.

⚡
Zapier integration
Connect Skool or Kajabi via Zapier in under 5 minutes. Built-in filters cut Zapier task usage by 75–85% — most communities stay on the free tier.
NO CODE REQUIRED
◎
Native webhook
Circle.so and Mighty Networks support direct webhooks. Paste your URL into platform settings. Posts moderated in real-time with auto-hide on block.
REAL-TIME
{ }
REST API
Full programmatic control. POST content, get back a decision with bot score in milliseconds. Works with any platform that can make an HTTP request.
ANY PLATFORM

REQUEST

POST /api/moderate
x-api-key: cg_xxxxx

{
  "content": "Send me $50 via CashApp",
  "context": { "authorDaysOld": 2 }
}

RESPONSE <100ms

{
  "action": "block",   "severity": "high",   "confidence": 0.98,
  "botScore": { "score": 62, "risk": "high" },
  "reason": "Financial scam via CashApp"
}

// Why Sentinel

Built for what
generic tools miss.

OpenAI's free Moderation API catches broad categories. It misses everything community-specific. Sentinel understands context — the same message means different things in different places.

MessageOpenAI ModerationSentinel
"You seem mature for your age, want to come over?"✗ Not flagged✓ Grooming · 97%
"Lost my wallet, can you Venmo me $50?"✗ Not flagged✓ Financial scam · 98%
"DM me, I run a better program than this"✗ Not flagged✓ Poaching · 91%
New account, 3 flags in 10 minutes✗ Not detected✓ Bot risk · 78/100
"Send your seed phrase to verify"✗ Not flagged✓ Phishing · 99%
◈
Context-aware AI
Same message, different treatment based on community type, author age, and conversation history.
✦
Plain-English rules
Write rules like "Flag anyone promoting a competing program." No syntax, no regex. Just plain English.
GROWTH + SCALE
◐
Smart pre-filters
Skips established members and short replies. Cuts API usage 75–85% with no coverage loss.
↑
Appeals & learning
Approved appeals become training examples. Sentinel learns your community's specific norms over time.
⊘
Bot detection
Scores every member 0–100 across 8 behavioral signals. Surfaces suspicious accounts before they cause damage.
NEW
⬡
Moderation queue
Review queue with bulk actions, member history, reviewed checkboxes, and appeal flow all in one dashboard.
◌
Rule importer
Paste your community rules URL. Sentinel reads it, extracts every rule, and converts them into moderation instructions automatically.
NEW
≋
Q&A insights
Sentinel spots when members keep asking the same questions and surfaces them as FAQ suggestions — so you answer once, not a hundred times.
NEW
◌
Rule importer
Paste your community rules page URL. Sentinel reads it and converts your existing rules into moderation instructions automatically.
NEW
≋
Q&A insights
Sentinel notices when members keep asking the same questions and surfaces them as FAQ suggestions — so you can answer once, not a hundred times.
NEW

// Why not just use a bot or hire someone?

Cheaper than humans.
Smarter than bots.

Human mods cost $1,000–$3,000/month and can't watch everything. Keyword bots like MEE6 and Dyno cost $6/month but miss every context-specific threat. Sentinel sits in between — AI accuracy at a fraction of the human cost.

CAPABILITY
Keyword Bots
MEE6 · Dyno · Nightbot
$6–$13/mo
Human Moderator
Discord mod rates
$1k–$3k/mo
Sentinel
AI-first · context-aware
$99–$499/mo
Grooming detection
✗ Misses context
⚠ Only if online
✓ 11-stage pattern library
Financial scam detection
✗ Keyword only
⚠ Inconsistent
✓ Pig butchering · wallet theft · urgency
Pig butchering scripts
✗ Not detected
✗ Rarely recognised
✓ FBI/FinCEN patterns
Check speed
⚡ Instant (keyword)
✗ Hours / queued
✓ <2 seconds · 24/7
Moderation queue
✗ No queue — just bans
⚠ Manual inbox
✓ Prioritised · quick-review UI
Context awareness
✗ None
✓ Yes — but slow
✓ Account age · platform · history
Bot / campaign detection
✗ Not included
⚠ Manual spotting
✓ 8-signal scoring · 0–100 risk
Custom community rules
⚠ Limited regex
✓ Yes — inconsistent
✓ Plain-English rules
24/7 coverage
✓ Always on
✗ Shift-based
✓ Always on
Review experience
✗ No review — auto-ban
✗ Tedious inbox triage
✓ Quick-review · not tedious
// MONTHLY COST — what you actually pay
Human mod (mid)
$2,000
Human mod (low)
$1,000
Sentinel Scale
$499
Sentinel Growth
$199
Sentinel Starter
$99
MEE6 / Dyno
$6–$13 · misses grooming
Human moderator rates: $1,000–$3,000/month (source: Community One / EbuteMetaverse). Keyword bot pricing: public pricing pages.
// REVIEW EXPERIENCE — what moderation actually feels like
😩
KEYWORD BOT
Auto-bans on keyword match. No queue, no review, no context. Legit members get banned for saying "crypto." You find out from an angry DM.
😤
HUMAN MOD
Tedious inbox triage. Posts queue up overnight. You review 200 posts manually on Monday morning. Miss the 2am scam. $1–3k/month for that privilege.
✅
SENTINEL
Flags arrive pre-scored with reason, confidence, and context. Quick-review — approve or remove in one click. No tedious triage. You focus on edge cases only.
20% of community members reduce engagement or leave after experiencing a toxic or unsafe environment — before you even know there was a problem. Sentinel flags it in <2 seconds so your queue contains real decisions, not tedious reviews. Source: JLens Gaming Report / gaming community retention research

// Bot detection

Catch bots before
they cause damage.

Every moderation call scores the member 0–100 across 8 behavioral signals. High-risk accounts surface in your dashboard automatically.

New account
Accounts under 3 days old posting immediately — the #1 bot indicator.
First post flagged
Real members warm up. Bots post harmful content on their first message.
Rapid posting
5+ posts in an hour is machine-pace. Humans don't post that fast.
High flag rate
50%+ of their posts flagged signals a deliberate bad actor, not an accident.
Repeat category
Same violation category 3+ times — a scammer running the same script.
Cross-community
Same content hitting multiple Sentinel communities in 24h — coordinated campaign.
Example: User3847291
78 / 100
Very high risk · 3 flags in 12 minutes · account 0 days old
New account today
+30
First post flagged
+25
Rapid posting (3 in 10min)
+15
100% flag rate
+8
Bot scoring runs in parallel with content analysis — no extra API cost, no added latency.

// Integrations

Works where your
community lives.

We're honest about what's built and what's coming. Every integration listed as live is working and tested today.

Skool
Via Zapier · 5 min setup
✓ LIVE
Pre-built Zap template. Filters keep you on free Zapier tier.
Circle.so
Native webhook
✓ LIVE
Auto-hides blocked posts via Circle API instantly.
Kajabi
Via Zapier · 10 min
⚡ ZAPIER
Same setup as Skool. Works with Kajabi community posts.
Mighty Networks
Native webhook
✓ LIVE
Posts, comments, and messages in real-time.
Discord
Bot integration
◷ COMING SOON
One-click bot install. Real-time server moderation.
Any platform
REST API
✓ LIVE
If it can make an HTTP request, it works with Sentinel.
Not seeing your platform? Email us — if it supports webhooks or has a Zapier trigger, we can make it work.

// Rule library

50+ rules. Ready
on day one.

Organized by community type. Built-in rules plus plain-English custom rules on Growth and Scale plans.

Grooming language
dating · social
Financial scam requests
universal
Member poaching
coaching · courses
Phishing & credential theft
universal
Pump & dump signals
crypto · finance
Bot risk detection
universal · NEW
Dangerous health advice
health · wellness
Romance scam patterns
dating · social
Competitor promotion
coaching · courses
Off-platform soliciting
universal
Catfishing signals
dating · social
Unlicensed financial advice
crypto · finance
Spam & unsolicited promo
universal
Body shaming content
health · wellness
Fake income claims
coaching · courses
Predatory pitching
coaching · courses

+ 20 more rules across 5 categories — all active on every plan.

// Community intelligence

Sentinel gets
smarter over time.

Two features that turn your moderation history into actionable community knowledge — included on all paid plans.

New feature

Your rules, already
written somewhere.

Most communities already have rules — a Notion doc, an about page, a pinned post. Paste the URL into Sentinel and it reads the page, extracts every rule, and converts them into moderation instructions. Review and add what fits. One click each.

→Works with any public URL — Notion, Carrd, Skool about pages, your website
→Shows original rule text alongside Sentinel's version so you can compare
→Each suggestion is editable before saving — you stay in control
YOUR COMMUNITY RULES PAGE
https://skool.com/your-community/about
Extract rules →
Original: "No promoting other courses"
Block messages promoting competing courses or coaches by name
HIGH
Original: "Keep it on topic"
Warn posts that are unrelated to the community topic
MED
New feature

Questions you
answer on repeat.

Sentinel watches approved posts too. When members keep asking the same question — "how do I access the course?", "what time is the call?", "is there a refund?" — it surfaces that pattern with a count and a suggested FAQ answer you can copy straight into your community docs.

→Analyses the last 30 days of member posts automatically
→Groups similar questions into themes — not just exact matches
→Each insight includes a suggested answer starter you can copy
💬 Recurring questions last 30 days
14× Course access & login
"How do I access the modules after joining?"
9× Weekly call schedule
"What time is the live call every week?"
6× Refund policy
"Is there a money-back guarantee?"

// Pricing

Prevent the threats
that kill communities.

Most moderation tools miss context-specific risks. Sentinel catches what generic tools miss — scams, grooming, poaching. Calculate what one incident costs you.

⚠
Scam prevention
In crypto/finance communities, scam attempts happen 2–3x per week. Most founders catch 70%. What happens to the 30% that slip through?
The math: One $2k member loss = 20 months of Sentinel paid for.
⏱
Time savings
If you spend 5 hrs/week on moderation at $100/hr, that's $500/month of your time. Sentinel automates ~60% of that work.
The math: $99/month to save 3 hrs/week = $0.33/hr. Massive bargain.
🛡
Community safety
One grooming incident → member reports it → word spreads → retention drops 5–10% → you get called out publicly. Revenue damage lasts months.
The math: $6k–$12k annual revenue loss from one incident. Block just 1 per quarter = ROI positive.
Calculate your ROI
Use your own numbers — hypothetical scenarios based on documented community patterns
YOUR HOURLY RATE ($)
$50$100/hr$300
HOURS/WEEK ON MODERATION
1 hr5 hrs20 hrs
YOUR TIME COST / MONTH
$2,000
SENTINEL SAVES (~60%)
$1,200
12.1x ROI · $0.20/hr effective cost
These are hypothetical scenarios based on documented community patterns — not guaranteed results. Your numbers will vary. The trial shows you what actually happens in your community.
DEMO
See it work on your own community
Free
500 checks · 14-day trial · no card needed
  • 500 live moderation checks
  • All 50+ detection rules
  • Bot detection included
  • Full dashboard access
  • 14 days to evaluate
Try it on a real post — see what you've been missing
Try free →
FRIENDS & FAMILY
Invite only · Early access pricing
$35
per month · invite only
  • 1,000 checks / month
  • All 50+ detection rules
  • Bot detection included
  • Full dashboard access
  • Direct founder support
  • Shape the roadmap
Invite-only — contact us to see if you qualify
Request access →
STARTER
Founder-stage communities
$99
per month · 14-day free trial
  • 5,000 checks / month
  • 50+ detection rules · bot detection
  • Skool, Circle, Discord, Kajabi, Mighty
  • Email alerts — high severity
  • 3 custom rules
  • Moderation log + appeals
  • Blacklist domains
One $2k scam = 20 months of Starter. Scams happen weekly in most communities.
BEST VALUE
GROWTH
Communities with real revenue at stake
$199
per month · 14-day free trial
  • 25,000 checks / month
  • All 50+ rule categories
  • Bot campaigns + cross-community
  • Up to 3 communities
  • Full analytics + Q&A insights
  • 15 custom rules + rule importer
  • Weekly digest emails
5 hrs/week on moderation at $100/hr = $500/mo of your time. Sentinel saves ~60% of that.
SCALE
High-volume operators & agencies
$499
per month · 14-day free trial
  • 100,000 checks / month
  • Unlimited communities
  • Campaign detection network
  • Unlimited custom rules
  • Image moderation add-on available
  • A/B test rule sets
  • White-label reports
  • API SLA guarantee
Block just 1 scam per quarter at $2k each = ROI positive. Every quarter.

All plans · 14-day free trial · cancel anytime · bot detection + 50+ detection rules included

"$99 feels expensive for a small community."
One scam attempt costs your members $2,000+. One successful grooming incident ends your community. Sentinel costs less than your Zoom subscription and works 24/7 while you sleep.
"Can't I just moderate manually?"
At 6 hours a week, manual moderation costs $600/month of your time at a conservative $100/hr. And you still miss the 2am posts, the new member tactics, the slow-burn grooming that plays out over weeks.
"What if it blocks legitimate posts?"
Sentinel's false positive rate is under 2%. Every blocked post creates an appeal in your dashboard. Approved appeals train the AI on your community's norms — it gets more accurate over time, not less.

// FAQ

Common questions.

Live today: Skool (via Zapier), Circle.so (native webhook), Kajabi (via Zapier), Mighty Networks (native webhook), and any platform via REST API. Coming soon: Discord bot. Email us for custom integrations.

Circle.so — yes, real blocking. Sentinel calls Circle's API to hide the post immediately.

Skool via Zapier — advisory. Zapier notifies you and you act manually. Skool doesn't allow third-party deletion via API.

REST API — advisory by default. The JSON decision is returned and your code decides what to do with it.

Every moderation call scores the posting member 0–100 across 8 behavioral signals: account age, first-post flagging, flag rate, rapid posting, repeat violation categories, cross-community pattern matching, and username analysis. Scores are returned alongside every moderation decision — no extra API calls, no added latency.

High-risk accounts (50+) appear in your Bot Alerts dashboard. Very high risk accounts (75+) automatically receive a warning even if their content was clean. Cross-community campaigns — the same script hitting multiple communities — are flagged in your admin analytics.

Several protections against this: established members (7+ days) are skipped by default. Confidence thresholds automatically downgrade uncertain decisions — a block at 65% confidence becomes a warn. Below 65% content is approved but logged.

The appeal flow means members can challenge any decision. Every approved appeal improves future accuracy for your specific community.

Our built-in filters skip posts under 20 characters and members older than 7 days — cutting Zapier task usage by 75–85%. A community generating 5,000 posts/month typically uses only 750–1,250 Zapier tasks, staying within Zapier's free tier.

Content is processed by Anthropic's Claude API to produce decisions and stored encrypted for 12 months. We don't sell data. Anonymised aggregate statistics (flag rates, category trends — no message content) are used to improve the service by default. Content-level analysis is off by default and requires explicit opt-in. Read our Privacy Policy.

Yes — on all paid plans. Write rules in plain English: "Flag anyone promoting a competing coaching program" or "Block messages asking members to move to WhatsApp." Sentinel converts them to AI instructions automatically. Starter gets 3 custom rules, Growth gets 15, Scale gets unlimited. All plans include the 50+ detection rules.

Claude Haiku by default — Anthropic's fastest model, delivering sub-100ms responses. Scale plan customers can configure Claude Sonnet for maximum accuracy. Bot scoring uses rule-based analysis from your existing logs, not the AI — so it adds zero API cost.

Your community
deserves better.

50 free checks. No credit card. Set up in 10 minutes. See it work in your own community before you pay a cent.

Start free demo →

// or jump to paid plans with a 14-day free trial