Phase 1C #3: Prove spec-functions-aware return-binding helper preservation (StmtListDirectInternalHelperAssignStepInterfaceWithInternals) - #2224
Conversation
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
Bugbot couldn't run - usage limit reachedBugbot is counted against Cursor usage for this user or team, and this run hit a usage or spend limit. A user or team admin can review and increase usage limits in the Cursor dashboard. (requestId: serverGenReqId_84eb144d-af37-459e-bca8-90cdca75e16d) |
OpenCodeReview first-pass review🔁 Incomplete — this run did not finish and will retry on the same commit. Do not count it as review coverage. OCR produced no JSON output OCR stderr highlightsOCR pilot metrics & packet coverageOCR pilot metrics
Pilot mode: advisory only. Codex Review remains the merge gate. |
Th0rgal
left a comment
There was a problem hiding this comment.
Codex current-head verdict for 4de363d: APPROVE. (GitHub prevents the authenticated PR-author identity from recording a formal APPROVED review.)
Reviewed the complete diff against verified parent 9e8bb82. The change is confined to Compiler/Proofs/IRGeneration/GenericInduction/Calls.lean (+79) and generated PrintAxioms.lean (+2/-1; registry count 6033→6034). DirectInternalHelperAssignHeadStepCatalogWithInternals correctly mirrors PR #2223 for Stmt.internalCallAssign, preserves prefix-derived StmtOccursAtScope indexing, and requires CompiledStmtStepWithHelpersAndHelperIRWithInternals. The list-interface theorem uses the expected induction/generalizing recursion, routes internalCallAssign to the catalog witness, rejects other constructors via simp [stmtTouchesDirectInternalHelperAssignSurface], and embeds the tail with StmtOccursAtScope.tail. No changed-source sorry, admit, or axiom escape tokens were introduced.
Non-source/generated rows are limited to the PrintAxioms theorem registration and generated total comment; both are consistent and purely additive. Existing remote-build evidence is sufficient. The lone OCR semantic review failure is classified as retryable LLM/API-key infrastructure, not a source finding. GraphQL reviewThreads was empty at review time. No blocking correctness findings.
|
@codex review Current-head verdict: APPROVE for The diff is confined to the writer milestone ( |
Th0rgal
left a comment
There was a problem hiding this comment.
@codex review
Exact-head Codex review for 4de363d4635758abe7cb766df53687aa68520f33 against base 9e8bb828309e4b924c45ac9b19bb56deef41063a.
VERDICT: CLEAN
The assign-only spec-functions-aware catalog uses CompiledStmtStepWithHelpersAndHelperIRWithInternals for Stmt.internalCallAssign, preserves prefix-derived StmtOccursAtScope indexing, and the list-interface proof mirrors the predecessor recursion and tail embedding. The diff is confined to Calls.lean (+79) and generated PrintAxioms.lean (+2/-1), including the 6033→6034 registration update. No new sorry, admit, or axiom occurs in changed source. No unresolved non-outdated technical review threads were found.
|
Codex Review: Something went wrong. Try again later by commenting “@codex review”. 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". |
Th0rgal
left a comment
There was a problem hiding this comment.
@codex review
COMMENTED (formal APPROVE unavailable because GitHub does not allow the PR author to approve their own PR) — clean current-head review at 4de363d4635758abe7cb766df53687aa68520f33 against parent 9e8bb828309e4b924c45ac9b19bb56deef41063a.
The diff is confined to the expected changes (Calls.lean +79; PrintAxioms.lean +2/-1). The assign-only WithInternals catalog uses CompiledStmtStepWithHelpersAndHelperIRWithInternals for Stmt.internalCallAssign, preserves prefix-derived StmtOccursAtScope indexing, and the list-interface recursion mirrors the predecessor call-only theorem with correct constructor routing and tail embedding. PrintAxioms.lean adds the expected theorem and updates 6033 to 6034. No new sorry, admit, or axiom occurs in changed source. GraphQL reports zero unresolved non-outdated review threads. No source-correctness findings.
|
Codex Review: Didn't find any major issues. Swish! 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". |
|
APPROVE |
Summary
Stmt.internalCallAssignoccurrences at their prefix-derived scopesStmtListDirectInternalHelperAssignStepInterfaceWithInternalsby the same scoped list recursion used by predecessor PR Phase 1C #2: Prove spec-functions-aware helper-call preservation (StmtListDirectInternalHelperCallStepInterfaceWithInternals) #2223Verification
lake build PrintAxioms(2569 jobs, exit 0; no cached terminal-job replay)lake env lean PrintAxioms.lean(exit 0)lake build(2228 jobs, exit 0)0 sorry'dtotal headerContinues PR #2223 and addresses issue #2080, Phase 1C case 3/4.
Note
Low Risk
Additive Lean proof scaffolding only; no runtime, auth, or compilation behavior changes.
Overview
Adds
DirectInternalHelperAssignHeadStepCatalogWithInternals, an assign-only catalog that holdsCompiledStmtStepWithHelpersAndHelperIRWithInternalswitnesses for eachStmt.internalCallAssignat its prefix-derived scope (viaStmtOccursAtScope), matching how those heads compile throughcompileInternalCallArgs ... spec.functionsrather than the empty-internals path in the legacy catalog.Proves
stmtListDirectInternalHelperAssignStepInterfaceWithInternals_of_headStepCatalog, which walks the function body with the same scoped list recursion as the void-callWithInternalstheorem from #2223 and lifts the catalog intoStmtListDirectInternalHelperAssignStepInterfaceWithInternalsfor the whole body.Registers the new public theorem in the generated
PrintAxiomsaudit (6034 total lemmas).Reviewed by Cursor Bugbot for commit 4de363d. Bugbot is set up for automated code reviews on this repo. Configure here.