⚠ PRELIMINARY BLOCKCHAIN TRACENot legal advice. Not fund recovery. Confidence grading reflects what is visibly inferable on-chain only.minimise
Worked example

From seed transaction to solicitor-ready questions

Vaultoro-style pilot trace

A sample claimant record starts with seed transaction c5ab64..., a 36 BTC input, and 29 outputs. Krystal Trace separates directly evidenced outputs from inferred change paths, flags common-endpoint behaviour, and records the confidence grade next to each claim.

Input36 BTC seed transactionOutputs29 visible branchesHeuristicChange-address candidate requires reviewExportMD + PDF evidence pack
Seed TXID36 BTCSplit cluster29 outputsExchange leadgrade BChange pathreview
ADirectly evidencedBStrong inferenceCPlausible inferenceDWeak — reviewEUnknown
Product flow

A compact matter workspace, not a recovery desk

Dashboard mock

Matter
NVSL
62 claimant records indexed

The dashboard keeps claimant counts, open traces, evidence gaps, and report status visible without live-data dependencies on marketing pages.

Trace view mock

Seed TXIDA
Exchange endpointB
Change addressC

Each edge in the trace keeps the confidence grade beside the narrative so a reviewer can see what is evidence and what is inference.

PDF output mock

Report cover
PRELIMINARY TRACE
Not legal advice
Not fund recovery

Sub-processors: Ollama Inc. / Anthropic PBC
Confidence grades

Worked examples of A-E labels

ADirectly evidenced

Directly evidenced on-chain, with a visible transaction path or source record.

BStrong inference

Strong inference from repeated on-chain behaviour or corroborating evidence.

CPlausible inference

Plausible inference that should be treated as a working lead.

DWeak — review

Weak signal requiring manual review before relying on it.

EUnknown

Unknown or not supportable from the evidence available.

Sub-processors/Ollama Inc. (US · transient · no training)/Anthropic PBC (US · GDPR DPA · no training)/Tier in use · public_marketingLocal-only mode available via sovereign flag