VerdictSwarm
ScanPortfolioPlansDocsFAQ

Token Intelligence

Scan before
you ape.

Six AI agents argue about your token so you don't get rugged.

Supports
SolanaSolana·BaseBase·VirtualsVirtuals

No signup. Free. Verdict in ~60 seconds.

60-second verdict·2,500+ tokens analyzed·20+ rug pulls caught

How it works

Four phases. One verdict.

01

Independent Analysis

Each agent inspects the token in parallel — contract, liquidity, sentiment, macro. No groupthink.

02

Adversarial Review

Devil's Advocate challenges every positive finding with counter-evidence.

03

Live Debate

When agents disagree by more than 2 points, they debate with evidence until they converge.

04

Consensus Verdict

The swarm produces a 0-100 score backed by every agent's reasoning and the debate trail.

Why this matters

One analyst misses things.
A swarm that disagrees catches them.

Most token scanners run a single model with a single prompt and call it a score. VerdictSwarm forces adversarial consensus — the verdict only stabilizes after agents challenge each other with evidence.

Live on-chain data

Contract state, liquidity depth, holder distribution — read at scan time, not cached.

Built-in skepticism

Devil's Advocate stress-tests the bull case. Every positive claim needs counter-evidence to survive.

Transparent reasoning

See the score breakdown, each agent's verdict, and the exact issues that moved the final number.

Repeatable

Re-scan whenever conditions change. Designed for pre-trade checks, not one-off audits.

swarm terminal
$ scan initiated for So11...1112 on Solana
[technician] pulling market structure, liquidity, holder distribution
[security] checking mint authority, freeze controls, LP state
[tokenomics] reviewing concentration, slippage, supply-side risks
[social] mapping sentiment quality, influencer signals
[macro] comparing against sector and market conditions
[devil] challenging bullish case — forcing evidence-backed rebuttals
[debate] agents reconciling disagreements before final scoring
$ consensus verdict generated — score: 72 / grade: B
_

Verified catch

$CRABGE scored 23/100 — flagged for active mint authority, extreme vol/liq ratio, and 83% holder concentration. Price crashed 83% within 24 hours.

See the output

What a scan looks like

Every scan produces a full intelligence report — score, grade, security checks, agent reasoning, and debate transcripts.

Sample Verdict$CRABGE · Solana
23/100FDanger
🔴 Mint authority: Active — can mint unlimited tokens
🔴 Vol/Liq ratio: 47:1 — wash trading signal
⚠️ Top 10 holders control 83% of supply
⚠️ Token age: <48 hours
View full report →

The difference

Free found 2 risks. Pro found 7.

One of those 5 hidden risks was the rug pull trigger.

Free scan2 agents
✓ Contract verified
✓ Liquidity present
🔒 Holder concentration
🔒 Bundle detection
🔒 Social sentiment
🔒 Macro context
🔒 Devil's Advocate
Pro scan6 agents
✓ Contract verified
✓ Liquidity present
✓ 83% holder concentration — whale risk
✓ Bundle detected — coordinated buy
✓ Low social signal — organic interest minimal
✓ Bear market headwinds ignored
✓ DA: bull case collapses under scrutiny

Plans

More agents. Deeper analysis.

🔒Launch pricing: 100 spots at $19.99/mo for life

Free

2 AI agents (Technician + Security) · Score + Grade + Quick Checks

Free
Scan Free

Pro

Popular

6 AI agents with dynamic expertise weighting · Bidirectional Devil's Advocate — stress-tests bull AND bear cases

$19.99/mo
100 of 100 launch spots — lock in for life
Upgrade

Pro+

Coming soon

10 specialized AI agents including Insider Forensics & Smart Money tracking · Live Swarm Monitoring — 24hr continuous token surveillance

$49.99/mo
Follow for Pro+ Launch
Full technical comparison →

Check your next trade.

Paste any contract address. Get a verdict in seconds.

A Sentien Labs Product · VerdictSwarm

DocsTermsPrivacyRefundsGitHubX

Analysis only — not financial advice. DYOR.

OPINION ONLY - NOT FINANCIAL ADVICE
VerdictSwarm provides automated risk assessments based on public on-chain data. All "Verdicts" and "Risk Ratings" are opinions generated by AI models, not statements of fact. A "Flagged" rating does not accuse anyone of crimes. A "Low Risk" rating does not guarantee safety. Conduct your own research. We accept no liability for losses.