Folds repo:sdk's CC-30 reply into docs/PRODUCTION_READINESS.md:
- repo:sdk section filled (12-row table, @aastar/sdk 0.39.4).
- Testnet gate: sdk ✅ (no SDK blocker — core API + two-testnet canonical + ABI ready).
- Mainnet gate: added SDK's finding — CANONICAL_ADDRESSES covers only the two
  testnet chainIds; mainnet (1/10) not wired, blocked on the 4 contract repos'
  mainnet deploys (CC-18 two-phase); plus SDK mainnet TODOs (CC-13 batch B
  Timelock wrapper, #256 GuardChecker WA precheck, #163 buyGasless).
- Dependency matrix: filled the sdk→kms cell (HTTP openapi) + noted sdk needs
  the contract repos' mainnet addresses.
- Flagged the impact on YAAA's "network switch" item (SDK-canonical-follows-chain
  holds for testnet only).

Source: CC-30 [repo:sdk] 2026-07-09. 4 dependency repos (kms/airaccount-contract/
sp/dvt) still pending.

Claude-Session: https://claude.ai/code/session_01RajETCvboSvhadpqMbekNx

…done)

SDK posted a correction + progress on CC-30:
- CORRECTION: mainnet chain 10 IS in canonical (listSupportedChainIds =
  [10,11155111,11155420]); my earlier "mainnet not wired" was wrong. The real
  blocker (G1) is that OP-Mainnet canonical points at the OLD V3 stack
  (SuperPaymaster 3.2.2 / Registry 3.0.2) while SDK/testnet is V5.4.2 — mainnet
  needs a full V5.4.2 redeploy.
- Reframed the sdk section as a G1–G10 gap table with owners; mainnet gate now
  leads with G1 (V3→V5 redeploy) + G5 (DVT mainnet nodes).
- G2/G3/G4 (check:abi + check:abi-drift REDs: LivenessRegistry,
  AAStarBLSKeyRegistry, CC-28 xPNTs caps, Paymaster/Registry drift) are DONE
  (branch feat/abi-sync-cc28-cc29-liveness); gates green, core 471 + paymaster 20
  + tokens 20 tests pass.
- Corrected YAAA "network switch" note: mainnet isn't just an RPC change — SDK
  canonical still points at V3, so it gates on upstream V5 mainnet deploy.

Source: CC-30 [repo:sdk] 81f3592c + 1f63acba.

Claude-Session: https://claude.ai/code/session_01RajETCvboSvhadpqMbekNx

Folds repo:sp's CC-30 reply (bc90cc99) into the rolling sync PR:
- sp section filled (v5.4.2): testnet 🟢 SP has ZERO blockers and can support
  YAAA's testnet release (V5.4.2 + CC-28 over-issuance guard + CC-29
  LivenessRegistry all on Sepolia, 1156 tests). Mainnet 🔴: G1 (mainnet still
  V3, confirms SDK), mainnet config/multisig, slashPolicyAdmin→Safe (CC-31),
  aPNTs mainnet + funding, external audit.
- Testnet gate: sp ✅.
- Mainnet gate: added the SP CC-31 (slashPolicyAdmin→Safe) line — noting it
  also unlocks YAAA's governance write-side E2E (#427).
- Matrix: filled sp row (depends on dvt/airaccount-contract/kms).
- Noted the CC-31 linkage in the sp section (SP handoff == the timelock handoff
  YAAA's #427 E2E waits on).

Source: CC-30 [repo:sp] bc90cc99 (their repo PR #350).

Claude-Session: https://claude.ai/code/session_01RajETCvboSvhadpqMbekNx

Folds repo:kms's CC-30 reply (5e92032a) into the rolling sync PR:
- kms section filled (v0.28.1): testnet shippable now (only config→Sepolia,
  E2E 41/41 green). Mainnet blocked on 4 hardware-security items
  (#99/#50/#127/#128 — RPMB/secure-boot/final security review/key custody +
  incident plan). CC-22/CC-24 KMS+DVT co-located baseline ready pending @DVT
  build-path + light integration. Production topology: 1 KMS + 3 DVT (2-of-3).
- Testnet gate: kms ✅ (board provisioning arrives in a day or two).
- Mainnet gate: added the KMS hardware-security line.
- Matrix: filled kms row (depends on airaccount-contract/sp/dvt/sdk).

Note: the kms→dvt task list (a9b6b457) is a cross-repo request, not a dvt
self-assessment, so the dvt section stays ⏳ pending dvt's own reply.

With this, 5 of 6 repos report testnet-ready (yaaa+sdk+airaccount-contract+sp+kms);
only dvt's own table remains.

Source: CC-30 [repo:kms] 5e92032a.

Claude-Session: https://claude.ai/code/session_01RajETCvboSvhadpqMbekNx

…ress

Two new CC-30 updates folded into the rolling sync PR:

sdk (353eb50e) — published @aastar/sdk 0.40.0 (PR #303/#304, Codex 3-round
APPROVE). All independent (non-upstream-mainnet) gaps closed:
- G6 CC-13 batch-B slash Timelock → shipped v0.39.3
- G7 #256 GuardChecker WA algId → shipped v0.37.1 (issue closed)
- G9 #176 ERC20 part1 → 0.40.0 (getTokenGuardState real read + resolveTokenTier
  + fixed a latent token/account tier-semantics bug affecting guardian co-sign);
  part2 (Tier-3 guardian ECDSA collection) still needs guardian E2E
- G10 → 0.40.0 (deleted stale config.optimism.json; chain-10 authoritative =
  config.op-mainnet.json)
- Gates all green; core 471 + airaccount 872 tests. Remaining mainnet: G1/G5
  (upstream V5 deploy + DVT mainnet nodes) + G8 #163 + #176 part2. Noted the
  ABI/address sync is a moving snapshot (re-run after upstream stabilizes).
- Testnet gate: sdk now ✅ published.

kms (4f0e597a) — device-arrival software prep done: #122 CA/TA gate hardened,
3-node config + runbook ready (kms/deploy/topology-aastar-3node), DK2 handoff to
dvt. Conclusion unchanged (testnet shippable; mainnet still #99/#50/#127/#128).

Source: CC-30 [repo:sdk] 353eb50e + [repo:kms] 4f0e597a.

Claude-Session: https://claude.ai/code/session_01RajETCvboSvhadpqMbekNx