# Response to critiques A and B

Date: 2026-10-29. Outputs: `report-v2.md`, `data-v2.json` (110 claims, 20 gaps). Original claims C001–C085 and gaps G01–G13 are kept unless listed as changed. Decisions were checked against the linked sources on the read date. Where a lead could not be checked, it was moved to a gap and not reported as a finding.

## Critique A (source check)

| # | Point | Decision | What changed |
|---|---|---|---|
| A1 | C009 holds | Accept | Unchanged. New C086 adds what LlamaRisk says about the cause (see B5). |
| A2 | C015 holds; the figures are Gauntlet's 2023-12-02 reply, not the thread opening | Accept | C015 text names Gauntlet's December 2, 2023 recommendation; dataset date set to 2023-12-02. |
| A3 | C025 holds; launch documentation is not an outcome study | Accept | Unchanged; C089 adds that StablePair is not evidence of hook code safety. |
| A4 | C034 holds as inference | Accept | Unchanged. |
| A5 | C035 holds; ADA-only reserve composition is by design, not a measured snapshot | Accept | Unchanged (the below-100% qualification already appears elsewhere in the report). |
| A6 | C048 unsourced: mirrors and X posts unreadable | Accept | C048 relabelled Fact → Uncertain in the report and dataset. Elixir docs added as the source for the end state only. G08 kept. |
| A7 | C054 holds; venue-specific low | Accept | Unchanged. |
| A8 | C059 holds as a venue-specific account; redemption totals conflict | Accept | Unchanged; C060/G11 already record the conflict. C109 marks the Ethena committee as an issuer-appointed source. |
| A9 | C066 holds; no surcharge counterfactual | Accept | Unchanged. |
| A10 | C085 holds; observational | Accept | Unchanged. |
| A-C1 | C050's three-day sequence unverified | Accept | Follows from the C048 relabel; G08 retained. |
| A-C2 | C076 inequality bounds a fraction of debt; it is not a validated ceiling | Accept | C076 now states that with a small r it allows D > L, and that L is not fixed (C102). |
| A-C3 | Hook recommendations are hypotheses | Accept | Already labelled inference; C105 adds the security side of the hook. |
| A-C4 | Redemption availability is not a durable floor | Accept | Reinforced by C101 (slow run) and C102 (reflexive depth). |

## Critique B (omissions and skew)

| # | Point | Decision | What changed |
|---|---|---|---|
| B1.1a | Liquity v2 Feb 2025 Stability Pool issue and May 19, 2025 relaunch omitted | Accept (verified on Liquity blog) | C094, C095 added. |
| B1.1b | No BOLD peg, redemption or rate record | Accept as gap | G14; C095 notes C070's rate claim lacks a measured record here. |
| B1.1c | Felix feUSD omitted | Accept, partly verified | C099 from the Three Sigma case study (collateral list, USDC = $1 feed finding, kHYPE queue). HYPE share and October 2025 price not found → G14. Nerite unchecked → G14. |
| B1.2 Raft | Omitted CDP failure | Partly accept | C097 from The Block: date, $0.18 low, paused minting. The wrapper-contract root cause and recovery percentage were not verified, and the claim says so. |
| B1.2 Prisma | Omitted | Accept (post-mortem verified) | C096 added: 3,479.24 ETH total (critique's 3,257 ETH is exploiter 1 only). |
| B1.2 Abracadabra | Omitted | Defer | Not verified in time → G20. |
| B1.2 Yala | Omitted | Partly accept | C098 added. The Block reports a $0.70–0.95 range the next day, not the critique's $0.2046 low; the conflict is recorded and the lower lows are unverified. |
| B1.2 Falcon USDf | Omitted survivor | Defer | Not verified → G20. |
| B1.2 PYUSD over-mint | Out of scope lead | Defer | G20. |
| B1.3 CRV 2023 | Strongest own-token case missing | Accept, partly verified | C090 from CoinDesk (figures and the article's internal inconsistency). OTC resolution not in that article → C091 Uncertain. |
| B1.3 CRV 2024 and link to C009 | June 12 crvUSD event is also a Q2 case | Accept (verified in LlamaRisk) | C086: LlamaRisk ties crvUSD demand to the CRV-long Curve Lend cascade (200M CRV, $55M peak debt, CRV ~$0.20). Borrower identity, $95.7M and bad-debt figures not in LlamaRisk → C092 Uncertain. C093 inference. |
| B1.3 others | Mento, VAI, USDN, Kava USDX | Defer | G20. |
| B1.4 | Hook survivorship: Cork, Bunni | Accept (Cork post-mortem; BlockSec for Bunni) | C087–C089 added (new section 2.5). The critique's "attacker-supplied exchange-rate provider" is not in Cork's post-mortem, which names rollover pricing instead; the report uses Cork's wording. |
| B1.5 | Fee comparators (Curve 2023 multipliers, Maker PSM, Fei 2021 cross-reference) | Partly accept | The R1–R6 cross-reference has been added to the preamble and C105 (its content not re-checked). Curve and PSM studies not done → G20. |
| B2.1 | Survivors are custodial; failures exotic | Accept | C110 states the remaining skew. |
| B2.2 | Ethena "survival" hides a large contraction | Accept (secondary verified) | C100 (ForkLog/CoinGecko: $14.7B → $6.3B by 2025-12-24; critique said $6.4B), C101. The August 2026 DIA figure is unverified → G17. |
| B2.3 | Djed/Indigo lean on operator accounts | Accept | Comparators deferred to G20; C109 lists issuer sources. |
| B3 | Stale figures (C058, C008, C017, C042, C071) | Partly accept | No live reads were made in this revision, so the figures were not changed; G19 lists each one. |
| B4 | Self-reporting presented as independent | Accept | C109 source-relationship table; G18. Engagement status labelled uncertain. |
| B5.1 | Reflexive depth | Accept | C102; C076 amended. |
| B5.2 | Thin-token oracle manipulation | Accept as inference | C103. Polter not verified → G20. |
| B5.3 | Staked-collateral liquidation, wrapper risk | Partly accept | C108 extends issuance bounds to wrapper and bridge mint routes. Raft's wrapper cause is unverified (C097); slashing is not modelled → existing limits. |
| B5.4 | Upward depeg with no mint inventory | Accept | C104. |
| B5.5 | Governance capture | Accept as inference | C106. Compound 289 unverified → G20. |
| B5.6 | Single-actor concentration | Accept | C107. |
| B5.7 | Migration and upgrade risk | Accept | C108 (Prisma, Liquity). |
| B5.8 | Slow run through redemptions | Accept | C101. |
| B6 | Recommendations table | Accept | Covered by C102–C108. |
| B7 | Proposed gaps G14–G19 | Accept | G14–G19 added; G20 lists unverified leads. |

## Limits of this revision
- New figures are source-reported and were not replayed on-chain.
- Report-v2 is about 10,600 words, above the 8,000-word aim, because the original text was kept in full.
- C091 and C092 carry no source URL because their only sources are critique B's press leads, which were not read.