fix(bb): regen Prover.toml fixtures and forge-fmt Debug.sol

[AztecBot]

Summary

Fix CI failure on merge-train/barretenberg (run 25476931781, log http://ci.aztec-labs.com/1778128976063647).

nargo execute --program-dir noir-projects/noir-protocol-circuits/crates/private-kernel-init failed with:

Type Array { length: 139, typ: Field } is expected to have length 139 but value Vec(...) has length 127

The committed Prover.toml fixtures embed serialized VKs / proofs whose length is sensitive to recent barretenberg constant changes on this train (notably HIDING_KERNEL_LOG_N 16→17, new MAX_APPS_PER_KERNEL / CHONK_MAX_NUM_CIRCUITS / BATCH_MERGE_PROOF_SIZE). This is the exact failure mode documented in barretenberg/cpp/CLAUDE.md (Prover.toml Fixtures, added in #23014).

Changes

• Regenerate 12 Prover.toml fixtures via:

  AZTEC_GENERATE_TEST_DATA=1 FAKE_PROOFS=1 yarn workspace @aztec/end-to-end test full.test
  

• forge fmt barretenberg/sol/test/utils/Debug.sol — the file was reformatted mid-build by barretenberg/sol/bootstrap.sh build_sol (forge fmt || true), which raised [bb-sol-tests] ERROR: Noticed changes to rebuild patterns during CI run: barretenberg/sol/test/utils/Debug.sol warnings and perturbed downstream cache hashes. Pre-formatting at rest eliminates the drift.

Verification

nargo execute on all 18 affected crates locally: Circuit witness successfully solved ✅ (12 regenerated + the 6 not-covered-by-full.test crates that turned out to still be valid, so the orchestrator_single_checkpoint regen step was not needed).

Detailed analysis: https://gist.github.com/AztecBot/293ccaae4f9337c64e4c5a5c4f44bee3

ClaudeBox log: https://claudebox.work/s/b956259ee219c45e?run=1

[AztecBot]

Automatically closing this stale claudebox draft PR (no updates for 5+ days). Re-open if still needed.