Commit graph

5 commits

Author SHA1 Message Date
ee52d50e20 PHASE 2 COMPLETE ON BETRUSTED: THE FULL POINT-LEVEL LIFT
(verify_accepts_iff_decompress, button-enforced)

Port of the risc0 decompress chain (identical v4 gen shape: same
instance naming, same RangeFrom+closure from_bytes extraction):

- source patch cee0e17: decompress step_2 negate-then-conditional-assign
  (the documented sqrt_ratio_i rewrite); extract.sh: decompress
  un-opaqued, re-extracted - the step_1/step_2 external axioms vanish
  from the template, decompress is transparent.
- Proofs/DecompressSpec.lean: dalek port, instance rename
  Shared0FieldElement51 -> SharedAFieldElement51.
- Proofs/FromBytesSpec.lean: the v4 closure-loader walk
  (range_from_index_spec + closure_call_spec; window/telescope math
  identical to dalek).
- Proofs/DecompressMain.lean: decompress_of_canonical (standard three)
  + verify_accepts_iff_decompress (corollary verbatim):

    accept  <=>  decompress(R) = [k]*(-A) + [s]*B   (as points).

check.sh: 4-tier Phase 3b; full-lift cone exactly [3 standard +
Signature + sha512_hash3 + to_bytes + Error + Error.new]. Full button
green fresh.

THE FULL POINT-LEVEL LIFT NOW HOLDS ON ALL FOUR PYRAMIDS.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-07-06 01:53:39 +02:00
9620cf5dd4 THE SIGNATURE APEX on the betrusted fork: verify_accepts_iff, button-enforced
Third pyramid capped. Identical shape to the risc0 fork (both are
sha2-0.10 stacks): the hash oracle is the single monomorphic
sha512_hash3(R, A, m) call, extraction runs --no-default-features.

- gen/CurveSig: extracted verify glue, definitionally welded to the
  proven CurveField model (every curve and scalar call resolves to a
  certified definition; only the hash and wire formats are opaque).
- Proofs/SigApexSpec.lean (unchanged from dalek): verify_loop_full with
  the standard three-axiom cone, and verify_accepts_iff — the verifier
  accepts IFF compress([s]B - [k]A) = R byte-for-byte.
- check.sh Phase 3b enforces the apex cone to be EXACTLY
  [propext, Classical.choice, Quot.sound, ed25519.Signature,
   verifying.sha512_hash3, ed25519.Signature.to_bytes,
   signature.error.Error, signature.error.Error.new].
- extract.sh gains the reproducible CurveSig stanza (same recipe
  verified byte-exact on the risc0 fork this session).

Full check.sh green: all standard certificates + the apex audit.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-07-04 22:49:19 +02:00
b408df3765 Merge scalar into CurveField: one type universe, serial-only backend
Replicates dalek's gen merge: extract.sh co-extracts the Scalar52 backend
and scalar::from_bytes_mod_order[_wide] into the CurveField model, pins
RUSTFLAGS --cfg curve25519_dalek_backend="serial" (SIMD dispatch compiles
out; get_selected_backend becomes a real definition), and upgrades
unwrap_u8 to the documented Choice-model definition. ScalarDenote repoints
by one import line; both buttons green over the merged gen.

Prepares the verify-glue integration (CurveSig) on this fork.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-07-04 22:20:26 +02:00
d6e879a764 Double-scalar-mul enters the verified model: vartime_double_base extracted transparently
extract.sh now opens crate::backend::serial::scalar_mul::vartime_double_base
(the other scalar_mul strategies stay opaque): non_adjacent_form (with its
loops), NafLookupTable5 (from/select), the curve-model helpers and
vartime_double_base::mul itself land in gen/CurveField - the same
namespace as the proven edwards operations, so the coming double-and-add
induction can consume EdDouble/EdAddProjNiels/EdConvert directly.
Zero sorries, zero external axioms (the pinned sources carry documented
compat refactors: single-assignment loop helpers, param-rooted while,
always-256-iterations, index-based LE load).

Full check.sh pressed fresh over the regenerated model: every existing
field and group-law certificate still green and axiom-clean - the scope
extension is purely additive.
2026-07-04 12:20:39 +02:00
bdafbf94bf lean-guard: disable core dumps (no more apport popups on capped aborts)
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-07-02 16:23:31 +02:00