Skip to content

Mind vNext: neutral challenge contract and exact verifier adapters - #51

Open
Kitahl wants to merge 1 commit into
mainfrom
upgrade/challenge-vnext-mind
Open

Mind vNext: neutral challenge contract and exact verifier adapters#51
Kitahl wants to merge 1 commit into
mainfrom
upgrade/challenge-vnext-mind

Conversation

@Kitahl

@Kitahl Kitahl commented Aug 27, 2026

Copy link
Copy Markdown
Owner

Scope

Implements the first skill slice from EGR-FOIL-INT-001 on the pinned baseline 4f088d688fa9e25b4608f44000a5d9812efa45f9:

  • WO-000 contract/authority freeze and import boundary
  • WO-001 additive egrt.challenge.v1 types, store, events, configuration, discriminator selection, and challenge gate
  • WO-002 generic candidate binding and distinct structural/semantic verification, with foil_authority.py compatibility wrappers
  • WO-003 Mind vNext native proof challenges and claim-native verifier receipts

Mind vNext behavior

  • candidate-bound alternate formalizations, negation, edge-case, and counterexample challenges
  • ordinal minimum-discriminator selection, capped at two load-bearing plans
  • exact symbolic-equivalence adapter with explicit assumptions and fail-closed availability
  • bounded exact enumeration/DP receipts
  • counterexample receipts
  • separate natural-language-to-formal scope status
  • explicit proof-bundle finalization

Boundaries

  • SHADOW remains the default challenge mode; OFF preserves baseline behavior
  • challenge resolutions never replace domain receipts
  • FOIL remains proposal-only and cannot clear proof obligations, execute repairs, or authorize writes
  • no non-FOIL module imports foil_*
  • no Soul, Space, Reality, Power, Time, Gauntlet-monitor, Meditate, Council, or FOIL-bridge upgrade is included
  • this implementation makes no behavioral-efficacy claim

Validation

All required pull-request workflows pass on clean squashed commit af8dc2621099dd4993272cbc9e623190ef23ca9c:

  • Research software validation: passed, including lint, runtime/layout tests, typed-runtime validation, showcase validation, syntax, and repository metadata
  • Security gates: passed, including full-history secret scan, dependency audits, deterministic hash-lock verification, and exact locked installation
  • Runtime portability: passed
  • CodeQL: passed
  • focused challenge/Mind implementation suite: 24 tests passed
  • complete repository unit run during implementation audit: 493 tests passed

The branch contains one implementation commit and no temporary payload or diagnostics workflow files.

@Kitahl
Kitahl force-pushed the upgrade/challenge-vnext-mind branch from 1d8e428 to 3c1d312 Compare August 27, 2026 00:30
@Kitahl
Kitahl force-pushed the upgrade/challenge-vnext-mind branch from 336a704 to af8dc26 Compare August 27, 2026 00:34
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant