CASEFILEPROVENANCE EXPLORER

FOREST SAVAGE · LEAD-REASSESSMENT PROTOTYPE

The evolving evidence pool

Follow a fictional search. Run the actual lead-review engine.
See which recorded decision needs a person’s attention, and why.

CASE 001 / FICTIONALMissing child · all names and events invented
CASE REPLAY01 / 07
CASE CLOCK09:00

STEP 01

PROVENANCE-DIRECTED REASSESSMENT

What the tool computes

The original Python engine follows declared links to request human review. People supply the case information.

Loading recorded output

Loading this step…

Awaiting review
—
Source versions retained
—
Decisions retained
—

The case story and the engine’s computed review requests are separate.

Compare a note-only edit or a missing link
Choose the engine input experiment

Use step 5 to compare the outcomes. A note-only edit changes source bytes without changing the mapped assessment. A missing link can prevent a request, so silence does not establish that a decision is current.

Inspect the engine record and source
Mapped fields changed
—
Recorded request path
—
Dependency coverage
—
Stored-source check
—

Copied-source hashes · Fictional inputs · Preserved engine source · Contribution and scope

Loading…

This demonstrates triage of decisions for review. The engine does not discover unrecorded relationships, rank investigative leads, or choose the outcome. Live Python loads on request from a pinned runtime CDN and runs in your browser.

THE RECORD AT THIS MOMENT

Connected evidence

FICTIONAL CASE
0Reports preserved
0Open search leads
0Known origins · open leads

Swipe the network to explore its connections. Every report also appears in the lead register below.

SourcesLead reportsDecisions
Recorded relationshipAssessment or revised basisChoose a node to trace it

Known origins describe the recorded source links. A different source does not by itself establish independent corroboration.

What this step reveals

PRESERVED THROUGH THIS STEP

Lead register

The story includes fictional human actions and assessments. Select any report to inspect its supplied source history.

ABOUT CASEFILE

A changed basis.
An explained request for review.

Press Next to follow the fictional case. The engine panel initially shows recorded output produced by the preserved Python engine. Press Run actual tool to execute those same Python files in this browser.

The source-link discovery and identification assessment are human-supplied case inputs. When a person corrects a separate identification-assessment record, the engine follows the declared relationship to the earlier decision and requests reassessment. The original witness report remains unchanged.

Every person, place, report, and outcome in the case is fictional. The browser worker computes review requests; it does not find the child, rank leads, discover shared origins, or make investigative decisions.

The note-only and missing-link controls are alternate experiments. They do not change the fixed story. Source links alone establish neither truth nor independence. The story JSON-LD and the actual Python ledger export are separately labeled.

This is a new demonstration of the later standalone ledger. It does not reproduce the paper’s earlier evaluation, validate CASE/UCO conformance, or establish a tamper-proof history. The copied engine retains its licenses and CaseLinker attribution.

Read the presenter walkthrough ↗