bridge: re-pin to the NIST-ACVP source commit; state the real coverage numbers

The companion fips205-source commit adds NIST ACVP SHA2-128s verification
vectors and a real differential bridge. This repo re-pins to it and replaces the
word "finite" with numbers, per external review rounds 4-6.

- extract.sh + PROVENANCE re-pinned 797b4ef -> 3153988. The provenance guard
  did its job first: it REFUSED the moved source until the pin was rotated
  deliberately.
- VERIFIED that the test/vector commit does not perturb the proved model: after
  re-extraction all four pinned model files are byte-identical
  (Types db720b4a…, Funs 7b7de55f…, TypesExternal 37958beb…, FunsExternal
  5efe551c…), check.sh is ALL GREEN, and the audit digest is unchanged
  (d83e297a…). The only regenerated difference is the untracked Aeneas
  *_Template.lean byproduct, which Phase 0 purges.
- TRUSTED-BASE item 9 and the README now state the bridge's actual size:
  131 assertion points (was 9), of which 20 are NIST ACVP SHA2-128s
  known-answer tests run against the proved path — 10 from the `internal`
  group (whose message IS M', exactly what slh_verify_128s consumes) and 10
  from `external pure` where mono, the deployed verifier and NIST must all
  agree, 9 of those with a NON-EMPTY context, which is the first empirical
  check of the domain-separator byte and context prefix that item 10 declares
  outside every proof. Both documents keep saying plainly that a passing
  differential test is evidence, not a proof.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

README.md

@@ -5,7 +5,7 @@ path**, extracted from a pure-Rust implementation into Lean 4 via
 Charon/Aeneas — the same pipeline, discipline, and honesty rules as the
 four ed25519 campaigns (`dalek/anza/risc0/betrusted-ed25519-verified`).
 
-## STATUS: eleven certificates over the extracted verify model (external review rounds 1–5 applied)
+## STATUS: eleven certificates over the extracted verify model (external review rounds 1–6 applied)
 
 `verification/check.sh` is **green** (exit 0): the model compiles, the proofs
 compile, and the audit passes. The audit runs **inside Lean**
@@ -64,8 +64,15 @@ cannot translate the deployed `Hashers` function-pointer struct):
   certificate — and it determines the FORS indices / WOTS digits, so a defect
   there could change the recomputed root while all eleven theorems still hold.
 - **Not the deployed public verifier:** the proved subject is the private
-  `verify_mono` facade; the bridge to upstream's generic `pk.verify()` is the
-  finite in-snapshot **differential test**, not a machine-checked refinement.
+  `verify_mono` facade; the bridge to upstream's generic `pk.verify()` is a
+  finite **differential test**, not a machine-checked refinement. Its size is
+  now stated rather than left to the word "finite": **131 assertion points**,
+  of which **20 are NIST ACVP SHA2-128s known-answer tests run against the
+  proved path** (see TRUSTED-BASE.md item 9). Until 2026-07-28 it was nine
+  points from a single seed, and this parameter set had *no* NIST verification
+  coverage at all — the vectors vendored upstream contain no SHA2-128s sigVer
+  group, so the 128s groups were extracted from the official NIST ACVP-Server
+  set (provenance recorded in the vector file).
 - **Not closed-form FIPS 205 correctness:** the folds are transliterations of
   the extracted loops (the hash primitives stay opaque); nothing here relates
   the recomputed root to a mathematical SLH-DSA specification.
@@ -156,8 +163,10 @@ casts, the WOTS+ checksum `iter().take()` + `&u32` Sub → an index loop —
 each site a local rewrite whose equivalence is argued in the commit and
 checked, for SHA2-128s, by the differential test; the obsoleted transpiler
 axioms were deleted from the external files); fidelity pinned by that
-differential test in the snapshot (valid / corrupted / wrong-message),
-re-run green after every source patch.
+differential test in the snapshot — since 2026-07-28 a randomized bridge
+(12 rounds, corruption across the whole signature, wrong-key and wrong-context
+cases) plus NIST ACVP 128s known-answer tests — re-run green after every
+source patch.
 
 - **`fips205.fors_inner_loop_eq`** + **`fips205.fors_outer_loop_eq`**
   (Algorithm 17, FORS pk-from-sig): a nested loop, split into two theorems.

TRUSTED-BASE.md

@@ -42,9 +42,25 @@ proceeds and is part of every claim.
    has no certificate; a defect there could change the recomputed root while
    all eleven theorems still hold.
 9. **The deployed generic verifier.** The proved subject is the private
-   `verify_mono` facade. The bridge to upstream's generic `pk.verify()` is
-   the finite in-snapshot differential test, not a machine-checked
-   refinement.
+   `verify_mono` facade. The bridge to upstream's generic `pk.verify()` is a
+   finite differential test, **not** a machine-checked refinement — no theorem
+   here says the two agree; the evidence is empirical and its size is stated so
+   a reader can judge it (external review, rounds 4–6, correctly objected that
+   "finite" without a number is not a disclosure):
+   - **131 assertion points** (was 9 until 2026-07-28: three rounds from one
+     fixed seed, corrupting one fixed byte of a 7856-byte signature);
+   - of those, **20 are NIST ACVP SHA2-128s known-answer tests run against the
+     proved path** — 10 from the `internal` group, whose message *is* M′ and so
+     is exactly what `slh_verify_128s` consumes, and 10 from the `external pure`
+     group where mono, the deployed verifier and NIST must all three agree.
+     NIST's negatives cover structurally distinct corruption sites (modified R,
+     SIGFORS, SIGHT, modified message) rather than one arbitrary byte;
+   - the remaining 108 are randomized: 12 rounds, varying message lengths
+     including empty, corruption spread across the whole signature, plus
+     wrong-public-key and wrong-context cases.
+   Still **not** covered by any of it: agreement on inputs nobody generated, and
+   the prehash variant against the mono path (see item 10). A passing
+   differential test is evidence, not a proof.
 10. **Everything above the extraction root.** The root is
     `verify_mono::slh_verify_128s = slh_verify_internal_free(M′, sig, pk)`,
     which takes the message-digest input **M′ as an argument**. The code in

verification/PROVENANCE.json

@@ -3,9 +3,9 @@
   "parameter_set": "SLH-DSA-SHA2-128s",
   "source": {
     "repo": "github.com/saymrwulf/fips205-source",
-    "head": "797b4ef26338e27363683656f93cb065a77daa0e",
+    "head": "3153988c4e89df66c41e698329f2ae5460880875",
     "upstream_base": "integritychain/fips205 @ 30bac08580aa61f653e5436d1bbacb5ffac446c4",
-    "upstream_deviation": "verbatim import minus CI workflows; then the additive monomorphic verify_mono compat module + de-plumbing rounds 1-2, each an individually-justified commit",
+    "upstream_deviation": "verbatim import minus CI workflows; then the additive monomorphic verify_mono compat module + de-plumbing rounds 1-2, each an individually-justified commit; then NIST ACVP SHA2-128s sigVer vectors + an expanded differential bridge (TEST-ONLY: no verify-path function changed, and re-running extract.sh against this commit reproduces all four model files byte-identically)",
     "cargo_lock": "committed",
     "rust_toolchain_toml": "nightly-2026-06-01"
   },
@@ -36,7 +36,7 @@
     "gen/SlhVerify/TypesExternal.lean",
     "gen/SlhVerify/FunsExternal.lean"
   ],
-  "byte_identical_regeneration_verified": "2026-07-24",
+  "byte_identical_regeneration_verified": "2026-07-28 (re-verified after the test/vector commit: all four pinned model files unchanged)",
   "model_integrity_sha256": {
     "_comment": "All four Lean model files check.sh Phase 0 pins before compiling anything \u2014 the two aeneas-generated files AND the two hand-maintained *External files (which carry the five oracle axioms and the Step defs). A hand-edit of the model fails the build here (round-4 reviewer F3).",
     "gen/SlhVerify/TypesExternal.lean": "37958beb7c90b649497a20b0802f98f0fb2d975c245eca956d7396af74d7759f",

verification/extract.sh

@@ -33,7 +33,7 @@ CRATE="${1:-$HOME/GitClone/FormalVerification/sources/fips205-source}"
 
 # The pinned source commit this repo's model + proofs were verified against.
 # Keep in lockstep with verification/PROVENANCE.json and the README snapshot.
-EXPECTED_SRC_COMMIT="${EXPECTED_SRC_COMMIT:-797b4ef26338e27363683656f93cb065a77daa0e}"
+EXPECTED_SRC_COMMIT="${EXPECTED_SRC_COMMIT:-3153988c4e89df66c41e698329f2ae5460880875}"
 
 # ── Provenance guard: refuse a wrong or dirty source tree (fail-closed) ──────
 [ -d "$CRATE/.git" ] || { echo "ERROR: '$CRATE' is not a git checkout of fips205-source." >&2; exit 2; }