lean

experiments/ — historical wiring records

This tree now contains the prose/audit record of retired integration witnesses. During the v13 custody cleanup, the superseded no_free_lift_wiring Lean project was deleted from live source after its ratified successor had long since become LeanProofs.Witnessed.*. The exact project remains recoverable from the v12 tag and Git history.

The remaining documents are provenance, not current build targets. Their historical green receipts do not promote a theorem, attest the current tree, or prove runtime conformance.

Current ownership is kept distinct:

tier what it is
LeanProofs/ public stable API and terminal public evidence
sibling skunkworks live formal incubation
experiments/ historical integration/audit prose; non-canonical

Historical custody contract

Custody-Class:  EXPERIMENTAL-WIRING
Build status:   observed in the archived tree; not live in v13
Citation tier:  non-authoritative integration witness
May cite:       module graph, axiom footprint, counterexamples, audit findings
May NOT cite:   doctrine ratification, runtime admission, canonical surface membership

The audit remains beside the ratification and migration records. Git history supplies the source to which those dated receipts refer.


no_free_lift_wiring/

The wired customs-office stack was the promotion provenance for the Witnessed Derivation Calculus. Its ratification and migration documents remain useful; its Lean source is no longer a second live implementation.

Two standing fences on interpreting the archive:

  1. The historical wiring was never runtime admission. Its proofs and build receipts do not establish that a consumer conforms.

  2. The atlas correspondence remains unverified. ATLAS-MAP.md maps this stack to ~/git/intake-composition-atlas (a real receipt-enforcing linter). That Rosetta is a candidate correspondence; whether the atlas actually exhibits the mapped behaviors (fixtures/fail/no-receipt.yaml, signed_is_not_witnessed, depth-1 cap) is not verified from this repo. Treat it as a historical map to chase, not a proven bridge.

New integration experiments should be opened only with an explicit current purpose and custody plan; live theorem incubation belongs in skunkworks rather than recreating a second Scratch/ lane here.