Fix multi-day shadow evidence and saved research recovery - #597
Merged
Conversation
Co-Authored-By: Codex <noreply@openai.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Multi-day shadow research could stop on its third observation because validating the adjacent predecessor recursively required an earlier record that the API could not receive. Validate each artifact's content separately from its adjacent evidence/receipt pair; consumers still traverse from the first record to establish the complete window. Serialization and content hashes do not establish ancestry or completion.
Saved research also rejected
needs_reviewandresearch_candidateproposals that the initial cycle had already accepted. Initial and resumed execution now share the existing allowed labels while retaining strict backtest, budget, identity and human-decision checks. Neither change grants live authority.Validation: regression cases failed before the fixes; 215 related tests passed with networking blocked, including third-day continuity, tampering, saved recovery, human accept/reject and terminal reuse. Ruff and diff checks passed. Implementation was independently inspected and the same bounded tests rerun. No real model, broker, market-data acquisition or deployment was exercised.