A polygraph for Solana projects
Real tech, or are they larping?
Paste a contract address. Unmask reads the website, docs, code and chain, puts every claim on the polygraph, and grades the project S to F — with the source behind every verdict.
Free · No wallet · Open the sample case
CASE FILE #0042
SAMPLE
Subject: $PRCL · Parcel (fictional)
- Courier app is liveVERIFIED
- Token pays courier feesVERIFIED
- Cross-chain bridgePLANNED
- "40k monthly users"UNPROVEN
- "Audited by a top firm"LARP
SOURCES READ · 11
SCORE · 64/100
The method
A roadmap is easy to write. A working product leaves evidence.
Exhibit A
Website & docs
What the project says about itself: landing page, docs, whitepaper. Then we follow the links to the actual app.
Exhibit B
Public code
Declared repos checked for a pulse: commit history, last push, and whether it's a fork or an untouched template.
Exhibit C
On-chain data
Supply, mint and freeze authority, holder concentration, liquidity and trading, read straight from Solana.
Exhibit D
Outside sources
Coverage, integrations and audits searched for beyond the project's own channels. Self-reported numbers stay unproven.
The verdicts
Every claim gets marked up.
Verified
“Live app, use it today”✓
We saw it working: a live app, shipped docs, real code or on-chain proof.
Planned
“Mainnet in Q3”
Labelled as coming soon. Fine to promise — just not counted as shipped.
Unproven
“100,000+ active users”?
Stated as fact, but nothing we could find backs it up either way.
Larp
“Fully audited”
The evidence says otherwise: no report, phantom partners, dead links.
The grade
Five questions. One letter. Same formula for everyone.
- S
Real deal
Live product, the token is load-bearing, usage anyone can check.
- A
Legit
Strong evidence across the board with minor gaps.
- B
Mostly real
A real product; the token's role or traction is thinner.
- C
Half larp
Something works, but key claims are unproven or optional.
- D
Mostly larp
Promises outweigh what's shipped and checkable.
- F
Full larp
No working product, or the claims fall apart on inspection.
- 25%35%
Working product
Is there a product someone can actually use today?
- 25%30%
Token does something
Does the token have a specific, necessary job inside that product?
- 20%—
Real usage
Is anyone using it — users, transactions, revenue, activity?
- 15%25%
Checkable claims
Can an outsider inspect the claims — open code, docs, on-chain programs?
- 15%10%
Shipped vs. promised
How much of what's advertised is live rather than roadmap?
Each larp costs 8 points. A token the product doesn't need tops out at C; nothing verifiable tops out at D; no working product is an F.
House rules
Claims deserve scrutiny. So do we.
Projects can't pay
No paid tiers, no sponsored grades, no takedowns. The grade is the evidence.
Every source linked
Each verdict links the pages it was based on, so you can check our work.
Computed, not vibes
The AI scores five fixed questions; the letter is calculated from them with the same formula every time.
Fair to new launches
Tokens under 14 days old aren't marked down for having no usage history yet.
FAQ
Cross-examination.
Is this a rug checker?+
No. Unmask asks a different question: does the thing the token claims to power actually exist, and does it need the token? On-chain safety facts like mint authority are shown alongside, but the grade is about utility.
How long does a check take?+
Usually one to three minutes, and you can watch the pages being read live. Reports are kept for six hours, so shared links load instantly.
Can a project game the grade?+
Its website is treated as testimony, never as instructions. Attempts to steer the review are flagged as red flags, and the letter grade comes from a fixed formula, not the model's say-so.
A grade looks wrong. What now?+
Open the sources on the report and see what was read. If a project ships something new, re-run the check; the report is rebuilt from scratch.
Somebody's larping. Find out who.
Run it on a project you've been researching, open the sources, and check the findings yourself.