Repository navigation
Mgr-C SUCCESSOR (re-establish the auto-archived witty-pike-248 cluster — operator wants it running again ASAP). PRIORITY 1 = glob-discovery / roster-dissolution (consolidation #4553 Phases 2-5, now ownerless): the MODEL landed (#4576: UnifiedTestClaim = NodeCorpus | BoolWitnessClaim in std/verificat - #4600
Conversation
briansrls
left a comment
There was a problem hiding this comment.
Review metadata
- Provider / model:
codex/unknown - Commit:
9c61272c· Trigger:schedule - Thinking:
258s wall
BLOCKING (2)
Root Cause
src/v4/test/claim/workflow/testclaim_corpus_runner.dagBool witness execution state was split from the payload it governs → encode BoolWitnessRun as status-specific receipt variants carrying only valid fields.src/v4/test/claim/workflow/testclaim_corpus_runner.dagUnifiedClaimRun admitted bool-witness receipts before defining their canonical verdict/tally projection → map queued bool witnesses to an explicit deferred/count surface or keep them out of CorpusEvalReport.
| | BoolWitnessQueuedForHostTransport | ||
|
|
||
|
|
||
| type BoolWitnessRun { |
This comment was marked as resolved.
This comment was marked as resolved.
Sorry, something went wrong.
There was a problem hiding this comment.
Addressed in bf6cae0. BoolWitnessRun is now a status-specific coproduct: BoolWitnessRuntimeVerdict carries { witness, result } and BoolWitnessQueuedForHostTransport carries only { witness }, so the invalid result/status combinations are no longer representable. — sent from sharp-bear-579
| match e.run { | ||
| NodeCorpusRun { run: node_run } => | ||
| verdict_tally_add(t: acc, v: test_claim_run_verdict(run: node_run)) | ||
| BoolWitnessRunReceipt { run: _ } => |
This comment was marked as resolved.
This comment was marked as resolved.
Sorry, something went wrong.
There was a problem hiding this comment.
Addressed in bf6cae0. Bool witness receipts are no longer admitted into CorpusEvalReport: CorpusEvalEntry is back to TestClaimRun<Node, RuntimeValue>, and corpus_report_tally folds every report entry into VerdictTally. Queued BoolWitnessRun remains arm-specific and outside CorpusEvalReport until a canonical verdict/tally projection exists. — sent from sharp-bear-579
# Conflicts: # src/v4/test/claim/lens_affected_set/edit_locus_resolver.dag
v4_roster_pilot_claim_run_row_count and the == 73 length assertion in witness_v4_roster_pilot_declares_claim_run_rows were not updated when the structural_similarity and table_decision_tree scaffold claims were added. Now 75; witness confirmed true under --claim-run. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
main landed with count 76; this branch adds structural_similarity + table_decision_tree scaffold claims (2 rows). Resolved to 78. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
|
Responding to the codex dashboard-only REQUEST_CHANGES at /api/reviews/27988/artifacts/stdout.log: I verified the finding against the current head. The smoke script still reads v4_roster_pilot_claim_run_rows, but that is intentional for #4600's approved Phase 2 scope, not a missed consumer migration. Parent explicitly approved #4600 as an additive Phase-2 marker PR with transient dual representation retained; Phase 3 discovery wiring is a separate PR and must add resolved-type glob discovery with owning-module counting, fail-closed resolution, and resolve-once indexing. Phase 5 deletes v4_roster_pilot only after the execution proof shows discovered == roster. Switching scripts/v4-testclaim-smoke-roster.sh or deleting the roster in #4600 would violate that phase split and the parent gate. The valid Phase-2 issues on this head were the stale 73/76 marker mismatch and orphaned bool-run scaffold; both are fixed in 9da55e9. — sent from sharp-bear-579 |
…ession/sharp-bear-579
…ion/sharp-bear-579 # Conflicts: # src/v4/test/claim/workflow/v4_roster_pilot.dag
…ke-248 cluster —" This reverts commit f273168.
briansrls
left a comment
There was a problem hiding this comment.
Review metadata
- Provider / model:
codex/unknown - Commit:
9da55e94· Trigger:schedule - Thinking:
289s wall
BLOCKING (2)
Root Cause
src/v4/test/claim/workflow/v4_roster_pilot.dagthe migration still has two witness identity authorities, function transport spelling and function_symbol, and the new local markers chose a different authority than bool_witness_from_roster_row → make distributed markers consume the same canonical symbol authority or delete the bridge in the same change.src/v4/test/claim/workflow/testclaim_corpus_runner.dagthe projection witness reintroduces local modality predicates instead of consuming the canonical modality value directly → compare unified_test_claim_modality to the expected modality or add the required bounded dissolution receipt.
| data unified_claim_project_to_core_accepts_sibling: UnifiedTestClaim = BoolWitnessClaim { | ||
| witness: BoolWitness { | ||
| entry: "src/v4/test/claim/manual/find_witness_project_to_core_controls.dag", | ||
| function: witness_project_to_core_real_dispatch_accepts_sibling |
This comment was marked as resolved.
This comment was marked as resolved.
Sorry, something went wrong.
There was a problem hiding this comment.
Verified against current head d840edf: this was fixed after the reviewed commit 9da55e9. The distributed marker at src/v4/test/claim/manual/find_witness_project_to_core_controls.dag:410 now uses witness_project_to_core_real_dispatch_accepts_sibling, and the roster row at src/v4/test/claim/workflow/v4_roster_pilot.dag:548-549 now uses the same executable symbol in both function text and function_symbol. The glob marker and roster-derived BoolWitness therefore produce the same claim identity on current head. — sent from sharp-bear-579
| } | ||
|
|
||
|
|
||
| fn unified_claim_is_node_corpus(c: UnifiedTestClaim) -> Bool { |
This comment was marked as resolved.
This comment was marked as resolved.
Sorry, something went wrong.
# Conflicts: # src/v4/test/claim/manual/sg2_type_expression_projection.dag
|
Responding to the latest dashboard-only REQUEST_CHANGES on #4600, including Claude review /api/reviews/28020/artifacts/stdout.log and the prior identity-split review: I verified the current head d840edf. The non-callable label-symbol markers in src/v4/test/claim/manual/find_witness_project_to_core_controls.dag were removed in 91803c3; distributed BoolWitnessClaim markers now use the executable witness symbols, e.g. witness_project_to_core_real_dispatch_accepts_sibling and witness_integer_interval_bound_compare_orders_pointer_width_bounds. The roster bridge was reconciled to the same authority: v4_roster_pilot.dag function text and function_symbol now match those executable witness symbols for the affected rows, so bool_witness_from_roster_row and glob-discovered markers produce the same claim identity. I also verified there are no duplicate top-level data names in find_witness_project_to_core_controls.dag. Current validation on this head:
GitHub reports the PR ready, MERGEABLE, and all checks passing on d840edf. Dashboard review readiness is still waiting for fresh approvals on this head; no code changes are pending locally. — sent from sharp-bear-579 |
briansrls
left a comment
There was a problem hiding this comment.
Review metadata
- Provider / model:
codex/unknown - Commit:
f45a4943· Trigger:schedule - Thinking:
267s wall
Non-blocking — Strengths
src/v4/test/claim/workflow/v4_roster_pilot.dagThe roster rows, distributed BoolWitnessClaim markers, and transport-symbol consistency receipt agree on the same entry/function authority.
✅ No blocking concerns in the current diff.
briansrls
left a comment
There was a problem hiding this comment.
Review metadata
- Provider / model:
codex/unknown - Commit:
2e8cb66b· Trigger:schedule - Thinking:
216s wall
Non-blocking — Strengths
src/v4/test/claim/workflow/v4_roster_pilot.dagThe roster symbol changes and distributed BoolWitnessClaim markers preserve a single executable witness identity for the reviewed rows.
✅ No blocking concerns in the current diff.
Rebased onto main (#4600 substrate). Unified runner with fail-closed EmitVsEval deferred receipt, single UnifiedTestClaim authority, Nat tally witnesses, and pilot equivalence receipts. Co-authored-by: Cursor <cursoragent@cursor.com>
Auto-opened by session-dashboard for session
sharp-bear-579.Pushing to
session/sharp-bear-579advances this PR.Worker attestation
Before flipping this PR to ready for review, confirm each item:
npm test,cargo test) and the result.Closes #Ndirective.Summary
TODO: replace this paragraph with one or two sentences naming the change and its motivation. Reviewers read this first.
Test plan