docs(e2e): TX-Value-Verification 5-doc for v5.4.0-beta.1 fresh redeploy + Codex 2-axis

[jhfnetboy]

What

The full TX-Value-Verification 5-document evidence layer for the v5.4.0-beta.1 fresh redeploy rehearsal (renames v5.4.0-final → v5.4.0-beta1-deploy — same release code, clean redeploy, not a new version). Companion to the merged rehearsal PR #296.

Doc	Content
01-TESTDATA	fresh addresses; version()=5.4.0 + APNTS_TOKEN==config.aPNTs (no migration)
02-PLAN	scenario↔TX matrix (9 Tier-A / 22 Tier-B / 11 ⛔ / 11 deferred)
03-RESULTS	37/37 effective; 90/90 TXs independently RPC-verified REAL status=1; GA blockers eth_call-verified
04-CAPABILITY-MAP	tx→capability→user-value (adversarial scoping)
05-CODEX-CHALLENGE	2-axis verdict (MET/PARTIAL/NOT-MET) + Codex resolution

Codex 2-axis adversarial review — honest downgrades applied

• AXIS-1 gap closed: added the 4 gasless-rerun fix-txs + version()/APNTS_TOKEN eth_call + git no-src-change evidence → 90/90 REAL, status=1, both GA blockers eth_call-confirmed.
• 🔴 KEY DOWNGRADE — the earlier "keep base credit 100, fine on OP" was overclaimed. Codex confirmed the H-1 credit ceiling does block balance-payers (no balance short-circuit in code), so on Sepolia L1 a fresh ENDUSER with xPNTs still can't be SP-sponsored (validate charge ~120 > 100 base tier). The "OP gas is cheaper so 100 is fine" reasoning is directionally sound but UNMEASURED → the OP base-credit value is an OPEN GA decision pending an actual OP gas measurement, not resolved.
• Owed (honest): agent handleOps, debt-accrual + repayDebt cycle, live ⛔ negatives.

Verdict

Deploy mechanics of v5.4.0-beta.1 faithfully validated (wiring, both GA blockers eth_call-confirmed, deploy/test-script gaps found+fixed in #296, 90/90 TXs REAL). Several business capabilities remain PARTIAL/owed; the OP base-credit-tier value is unresolved pending an OP gas measurement.

Methodology: Dev-QualityControlFlow/TX-Value-Verification/. Mirrors docs/e2e/v5.4.0-beta.1/.

[chatgpt-codex-connector]

You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard.

[clestons]

Review: docs(e2e) TX-Value-Verification 5-doc for v5.4.0-beta.1 fresh redeploy — APPROVE [2-round]

Docs-only evidence layer (+608/-0, 6 files), companion to the merged rehearsal #296. Reviewed for honesty/no-overstatement and factual accuracy of the headline claims.

Verified

• Sampled TXs are genuinely REAL status=0x1 — spot-checked 4 of the cited hashes via Sepolia RPC: all status=0x1, blocks 11071169–11071273 (consistent with the #296 rehearsal window). The "90/90 RPC-verified REAL" claim holds for the sample.
• GA-blocker claims align with #296 — APNTS_TOKEN == config.aPNTs (0x9e66B457…, which I independently confirmed is the "AAStar PNTs" token in #296) and version()=5.4.0, both eth_call-verified per the doc.
• Exemplary honest scoping — this is the rare evidence PR that downgrades its own prior claims: the earlier "keep base credit 100, fine on OP" is explicitly walked back (Codex confirmed the H-1 credit ceiling blocks balance-payers — no balance short-circuit in code), and the OP base-credit value is correctly marked an OPEN GA decision pending an actual OP gas measurement, not resolved. handleOps/repayDebt/live-negatives openly listed as owed. No overstatement.
• Rename v5.4.0-final → v5.4.0-beta1-deploy (same release code, clean redeploy — not a new version) is consistent with #296.

Docs only, claims verified and honestly bounded. Verdict: APPROVE. The unresolved OP base-credit-tier decision is correctly surfaced for you, not buried.