# Evidence Task Engine

Updated: 2026-07-02

Boundary: This is an investigation workflow, not an accusation system. A task records what public source to pull next, what the source can prove, what it cannot prove, and how the result changes confidence.

## Statuses

- open
- pulled
- verified
- rejected
- needs counter-source
- privacy blocked

## Confidence Grades

- A: confirmed by primary public record, court record, official disclosure, regulatory filing, or authenticated exhibit
- B: strong lead supported by reputable reporting or multiple records but missing a primary record
- C: open lead requiring verification
- X: rejected, falsified, or speculation quarantined

## Current Tasks

1. EFT-001 — EFTA01104262 capital-class map: verify every visible node, label and relationship line against the original PDF.
2. EFT-002 — Billionaire / wealth-node cards: add dated net-worth source, filings, foundation records, political-money records and counter-source.
3. EFT-003 — Institution sector cards: pull record type, action, public statement, enforcement or review record.
4. EFT-004 — Money route cards: attach best source route and confidence rule to each money lane.
5. EFT-005 — Aviation / historical transport evidence: use historical exhibits, registry context and privacy policy only.
6. EFT-006 — Dark Speculation Lab claims: attach best proving source and best counter-source.
7. EFT-007 — Source Document Vault routing: make the Top 50 Evidence Pull List a required route for tracker updates.
8. EFT-008 — Public denial / response lane: archive statements, denials, corrections and responses.
9. EFT-009 — Evidence task status dashboard: count open, pulled, verified, rejected, counter-source and privacy-blocked tasks.
10. EFT-010 — Counter-source requirement: find correction, court rejection, misidentification, fake-document warning or reputable debunk.

Use `data/evidence-task-engine.json` for the full structured task cards.