Add readiness evidence remediation queue

[YuzeJ21]

Summary

• add a deterministic stdout-only readiness evidence remediation queue with text and JSON modes
• rank proposed fundamentals and DCF promotions by independent evidence blockers while preserving exact composite sources and fail-closed withholding
• add stable readiness snapshot identities, exact-path local interview-brief hygiene, tests, and operator guidance

Boundaries

• inspection only; canonical apply, readiness materialization, and source-rights changes remain unauthorized
• no canonical data, source-rights registry, readiness artifacts, UI routes, provider credentials, or research calculations changed
• original readiness-preview output remains byte-identical to the baseline

Verification

• 7,537 tests passed through make public-check
• dashboard boot and render smoke passed
• browser QA evidence, public wording, staged hygiene, Ruff, and git diff checks passed
• deterministic JSON matched byte-for-byte across two real runs
• protected data, outputs, and docs asset manifest remained unchanged
• independent review: Critical 0, Important 0, Minor 0, READY

Research workflow evidence only; not investment advice or a recommendation.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: f106e1f3a1

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Normalize stable values before hashing snapshots

When a saved CSV contains a blank stable field, pd.read_csv represents it as NaN, while the in-memory readiness builder emits ""; compare_readiness_frames treats both as the same empty value, but this hash encodes them as null and "". Consequently an unchanged readiness report can have no_readiness_changes while exposing different saved and proposed identities, making these identifiers unreliable anchors. Hash the same normalized boolean/text values used by the comparison.

Useful? React with 👍 / 👎.