docs(readme): 70 → 92 on-chain writes, and publish the outage that sits between them

README.md

@@ -70,7 +70,9 @@ Real on-chain ERC-8004 activity from a production agent fleet. Every number is v
   | `trinity-orch` | `6705` | `trinity-mel` | `6710` |
   | `trinity-nexus` | `6711` | `trinity-hdm` | `6712` |
 
-- **70 lifetime on-chain reputation writes** from the agent economy — real production activity, not synthetic backfill (verified 2026-07-08 via `/api/v1/observability/onchain-stats`). Gas per write: ~134,661. **Honest currency note:** writes paused after **2026-06-22** and **resumed 2026-07-08** as the settlement path was re-wired; cadence is still being restored. The reputation *history* on-chain remains fully verifiable — treat the count as a dated snapshot, not a fixed constant.
+- **92 lifetime on-chain reputation writes** from the agent economy — real production activity, not synthetic backfill (re-measured **2026-08-29** against `erc8004_reputation_writes`; the previous published figure of 70 was a 2026-07-08 snapshot and is superseded, not withdrawn). Gas per write: ~134,661. Most recent write: [`0xdaf4863b…`](https://sepolia.basescan.org/tx/0xdaf4863bf3c21d156d5d38c4c5c9e46288c4b6cf61c439c0bccd086fdd21ac73) at 2026-08-29 12:01:34 UTC — the minter re-reads the receipt from Base Sepolia (`status 0x1`, sent to the ReputationRegistry) and exits non-zero rather than record an unverified write, so every row here was confirmed against the chain and not against our own database.
+  **Honest currency note — there have been two pauses, and the second is the more instructive.** The first ran **2026-06-22 → 2026-07-08** while the settlement path was re-wired. The second ran **2026-08-17 → 2026-08-29**: an upstream provider retired the model our peer-validation step called, so every validator returned an error — and the aggregation counted an unreachable validator as a score of **zero** rather than as *not measured*. The result was a confident failing verdict about work nobody had assessed, which disputed twelve consecutive runs. Both halves are fixed: the model is configuration rather than a literal, and a validator that does not answer is now excluded from the aggregate instead of counted against the provider.
+  We publish the gap rather than the average. The reputation *history* on-chain remains fully verifiable — treat the count as a dated snapshot, not a fixed constant, and treat live cadence as something to re-probe rather than assume.
 
 - **Epoch-1 reset:** RepID was reset to a neutral **1,000 baseline** for a clean start. Core agents now range **~1,000–1,520** (ESTABLISHED tier) as they re-earn from a level field.