Documentation
Detection & Scoring System
SpotBlock uses two complementary engines: a classic scanner + quality workflow and a Super-Analyst AI pipeline. Each section shows the essentials up front — open Details for the full mechanics.
Reliability & Governance Controls
Published entries are explainable, auditable, and conservatively gated: strict promotion criteria, end-to-end provenance, and separation of duties between detection, enrichment, and approval.
Governance model in full: the three controlled stages, versioned scanners, and reproducible analyses.
1) Scam Scanner (1st detection layer)
The front line: fast, signal-driven, optimized for continuous intake. It assigns a 0–100 confidence score from bytecode + token heuristics.
Signal scoring, the full label mapping (including the phishing special case), and provenance capture.
2) Super-Analyst + 🧩 Vulnerability Scanner (2nd layer)
Second layer: Super-Analyst AI plus a dedicated vulnerability scanner — deeper contract-level analysis before promotion and review decisions.
🚀 Promotion output
Super-Analyst returns structured JSON including a rating in safe | suspicious | threat | unknown. Only suspicious / threat results are promoted into production entries for public explorer display; safe / unknown stays in history for traceability.
Retrieval layer (verified vs unverified retrievers) and analysis layer (AI mode, vuln mode, depths, dedup).
3) Publication Policy
Only high-confidence scanner detections are published. Medium-confidence hits stay in staging so they are not shown as scam or malicious in Explorer or Threat Check.
Full publication rules, including how on-demand Super-Analyst results are treated and how historical review comments are handled.
4) Compliance-Friendly Data Usage
Built so institutions can use the underlying data safely: each entry is traceable, evidence-backed, and conservatively published.
Evidence and provenance fields, traceable lifecycle storage, and the conservative publication rule.
5) User Scan Experience
Simple for users, with backend safeguards against redundant runs.
- Search address in explorer.
- If not found, connect wallet and sign scan request.
- Request enters queue and worker processes retriever + analyzer.
- If promoted, user is redirected to the full explorer card with details.
- Re-run for same depth/version is blocked; rerun becomes available after analyzer version updates.
- Super-Analyst (Deep AI) runs typically finish in on the order of ~15 seconds but can take longer under load or for heavy contracts.