Repository navigation
Phase 6C: bind legacy evidence components before accepting summaries - #91
Conversation
Reviewer's GuideThe PR hardens Phase 6C legacy evidence validation by cryptographically binding retained MIDI receipts and representatives, sharing raw IT/audio rejection across upload and closeout, and tying historical donor source metadata to the committed workflow head while explicitly leaving independent donor rebuild provenance unresolved. Sequence diagram for Phase 6C evidence validationsequenceDiagram
participant Workflow as MIDI workflow
participant Corpus as phase6c-midi-corpus.py
participant Summary as Donor summary
participant Receipt as Retained observation
participant Candidate as Candidate representative
participant Git as Committed workflow head
Workflow->>Corpus: audit_public_tree
Corpus-->>Workflow: Reject raw IT/audio or structural MIDI payloads
Workflow->>Summary: donor_summary
Summary->>Receipt: sha256_file
Receipt-->>Summary: Relative path and SHA-256 binding
Summary->>Git: validate_donor_source_identity
Git-->>Summary: source_revision and committed itmodule2.c bytes
Summary->>Receipt: require_support
Receipt-->>Summary: Recheck identity and execution fields
Summary->>Candidate: Match ID, name, hash, and size
Candidate-->>Summary: Exactly one donor corpus member
Summary-->>Workflow: Validated evidence or fail closed
File-Level Changes
Tips and commandsInteracting with Sourcery
Customizing Your ExperienceAccess your dashboard to:
Getting Help
|
|
@codex Please Review this exact SHA de22ace against the existing contract. Report only actionable correctness defects, with a minimal reproduction, expected versus actual behavior, and affected lines. State whether each reproduction was executed or statically inferred. Don’t repeat fixed findings without a new failing case. Keep architectural suggestions separate and non-blocking. |
|
@sourcery-ai review Please assess exact head de22ace for actionable correctness defects in the current evidence-integrity contract. Keep the documented private-observation holds and unresolved donor rebuild-provenance boundary intact. |
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
|
Codex Review: Didn't find any major issues. Keep it up! Reviewed commit: ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
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". |
|
Sorry @EmergentMonk, you've used your own review budget of 250,000 diff characters for the last 7 days. You can request another review in 8 hours and 50 minutes by commenting |
Scope
Harden the existing legacy observation lane while both private-observation gates remain pending. This does not expand the held Phase 6C ladder or begin Phase 7.
Verification
Passed locally using ordinary Python tests and project-owned inert metadata/header markers:
Independent prepublication review found no blocking issues. Existing CI path filters cover all changed files. Hosted checks and exact-final-head review remain required before merge.
Preserved boundaries
ROADMAP, parity classifications, closeout ledger, frozen manifests/fixtures/evidence, and both native source trees are unchanged. Both observation gates remain PENDING_PRIVATE_DISPATCH, the hold remains AWAITING_OBSERVATION_REVIEW, and parity remains UNKNOWN. No private corpus was processed and no historical evidence was backfilled. Older MIDI artifacts without the new bindings fail closed and require a fresh workflow run.
Historical donor probe hashes and ELF magic establish retained-artifact consistency only. A self-consistent substituted ELF is still outside the proof supplied here; independent donor rebuild provenance remains unresolved. The new tests do not claim native execution, compatibility, or private-observation completion.
Summary by Sourcery
Strengthen legacy Phase 6C evidence validation without advancing the observation gates or claiming private-observation completion.
Bug Fixes:
Enhancements:
Documentation:
Tests: