GOTRA Methodology

GOTRA Public Ledger is an auditable AI stock-research public ledger. It records public-safe demo predictions, resolved outcomes, visible errors, and research boundaries. It is research information only, not investment advice, not a trading signal, and not performance proof.

GOTRA Public Ledger 是一个可审计的 AI 股票研究公开预测账本，展示 public-safe demo 预测记录、已结算结果、公开错误和研究边界。它不是投资建议、不是交易信号、不是业绩证明。

How GOTRA records research objects

Each public record keeps a prediction identifier, ticker, company, sector, prediction date, horizon, direction, confidence, public-safe evidence labels, and outcome fields only when public-safe resolution data exists.

What public-safe demo data means

Public-safe demo data means this repository uses committed public JSON and public documentation sources only. It excludes raw provider/model I/O, prompts, completions, scorer transcripts, private run logs, databases, secrets, and local GOTRA experiment artifacts.

Resolved-only measurement

Resolved-only means only the 48 rows with public-safe numeric outcome and error fields are counted in resolved summaries. The 6 pending rows and 240 frozen-pending rows stay visible but are not counted as resolved outcomes.

Why pending rows are excluded

Pending rows do not have public-safe outcome fields in this repository build. Excluding them prevents fabricated backfills and keeps the audit boundary readable.

Why errors remain visible

Visible errors make the ledger auditable. A public ledger that hides misses would be weaker for review, attribution, and boundary checking.

What GOTRA does not claim

GOTRA Public Ledger provides public-safe research information only. It is not investment advice, not a trading signal, not live trading, not performance proof, and not a guarantee of future outcomes.