The Grading Process
Three licensed verifiers are drawn randomly from the verifier pool and assigned to your submission. Each verifier reviews independently — they cannot see each other’s scores. The median grade of the three verdicts is your final grade. Verifiers see:- Your submission and the evidence you provided
- The niche and tag type you assigned
- Belt activity on the submission (flags, confirms, fix attempts)
- The full knowledge tree context for the relevant niche
Verifiers earn 50 for a bad pass (approving bad intel) or a wrong reject (rejecting good intel). Their incentives are calibrated to reward careful, independent assessment. Becoming a verifier requires a $1,000 aSLEUTH balance — the Tier 2 threshold.
The Grade Table
Grades run from 1 to 10. The payout curve is superlinear — a Grade 10 is worth 25 times a Grade 1, not 10 times.The superlinear curve is intentional. Mediocre intel has low marginal value to the platform. Cross-domain, evidence-rich, novel intel is disproportionately more valuable — and the grade table reflects that.
Machine-Verified Intel
If your submission passes through the machine lane (on-chain-verifiable intel like wallet tags and contract labels), it does not go through verifier grading. Instead, it pays a fixed formula tariff set per category. The tariff is determined by the category’s standard value to the platform — you’ll see the applicable rate in the submission form before you submit.Retroactive Regrading
Accepted intel can be regraded after the fact. This works in both directions:- Upgrade: if your grade is revised upward, you receive the difference in payout immediately.
- Downgrade: if your grade is revised downward, there is no clawback. You keep what you were paid.
Retroactive upgrades are the platform’s way of rewarding intel that becomes more valuable over time — for example, a wallet tag that later gets confirmed by a major on-chain event. Check your contribution history periodically.