┌─ CASE FILE ───────────────────────────────────────────────┐
│  SUBJECT : Victor Nwachukwu   ($torm)                      │
│  TRADE   : on-chain forensics · behavioral risk            │
│  BASE    : Port Harcourt, NG                               │
│  STATUS  : Surveillance                                    │
└───────────────────────────────────────────────────────────┘

The ledger is a crime scene. I'm solving the case.

Everyone else reads the price. I read the intent.

I work the seam where raw chain data turns into evidence — behavioral risk scoring, token forensics, deployer reputation, entity clustering. The part most tools skip because it's hard: not what a token is, but what it's about to do to you.

Structure lies. On Solana every SPL token looks identical, so the tell isn't in the config — it's in the behavior. Liquidity that breathes wrong. A deployer who's done this before. Volume that's talking to itself. I built an engine to catch all of it, and I'm teaching it to follow money across every chain it runs to.

One rule I don't break: be the honest broker in a dirty room. Calibrated findings, evidence over assertion, confidence labeled, never slop.

RugBurn — a Solana-native behavioral risk and token-forensics engine. It doesn't ask how a token is built; it watches how it moves, then hands you a verdict you can defend.

behavior > structure        the discriminating signal lives in flow, not config
evidence-grade              append-only, hash-chained event log — every verdict auditable
follows the money           adapter pattern over a normalized transfer/entity graph,
                            so a trail doesn't die at the chain border
proven in the open          2x top-5, BirdEye On-Chain Intelligence hackathon —
                            10/10 across nearly every technical category

Now in the lab: carving RugBurn into a chain-agnostic forensics core — per-chain adapters feeding one normalized graph, so clustering and risk light up on any chain the moment its adapter lands. Solana was the hardest one. The rest compound.

• solana-compliance-skill — an agentic compliance & risk-intelligence team as a Claude Code skill. Token and wallet screening, regulatory gating (securities, AML, Travel Rule, US/EU/NG regimes), and forensic investigation — every output calibrated and held under human-in-the-loop oversight. Compliance that isn't slop.

• Stellar / Soroban contributions — merged contract and backend work across the ecosystem: partial-unstake logic with typed errors, price-oracle test coverage that surfaced a live latent bug, event-indexing, plus build/CI repairs that unblocked whole repos.

• The casebook — exploit post-mortems and rug-pattern breakdowns, written mechanism-first in a forensic voice. The reasoning is the product.

forensics / data    behavioral scoring · entity clustering · transfer-graph
                    normalization · deployer reputation · lifecycle staging
chains              Solana (SPL · Anchor) · Stellar / Soroban · EVM (adapter inbound)
languages           TypeScript · Go · Rust (active) · SQL
build               React · Next.js · Node · append-only hash-chained evidence stores
data layer          PostgreSQL · event-driven invalidation · staleness-aware caching ·
                    MCP delivery for agent + dApp consumption
sensibility         dark / minimal UI · perfectionist detail · ships production, not demos

I think in systems and I sweat the details — the forensics rigor and the design rigor come from the same place.

On-chain intelligence & forensics roles · grant-funded ecosystem work (SCF and kin) · collaborators serious about building risk infrastructure that actually protects people.

If you're chasing money across chains, hardening a protocol, or funding the tooling that keeps the space honest — my line's open.