How to read a claim without overstating its evidence

A matching endpoint is not a payment ledger

The 7 and 15 October account comparison joins debt quantities and accepted prices before interpreting the dollar total. A change in units valued at old prices, plus a change in prices applied to later units, reproduces the book change with a separately stated rounding residual. The residual is not allocated to fees or forgiveness. Exact endpoint decomposition.

The 15 October, 06:13:35 UTC scaled/index comparison is pointwise financial consistency. It does not certify every deployed rounding path or prove no offsetting borrowing and repayment. Equal cash balances can conceal gross arrivals and payments as well. A complete return calculation needs attributable cash, claims and expenses, not simply marked collateral minus marked debt. Indices, units and unknown gross flows.

Coverage must keep its sign and scope. Successful empty ordinary-account queries cover only the stated 3,170 October 7 blocks and topic position. No successful event coverage establishes the later 54,378-block interval. The later state block is 26,196,047; higher acquisition heads do not refresh its financial timestamp. A current block-hash witness does not fill a different earlier state’s missing hash or establish independent finality. Observation clocks and sources.

A reserve’s held cash, aToken claims, variable debt and unreconciled accounting remainder are different objects. A proportion such as cash/claims must name the token, endpoint, denominator and date. It is not the borrower’s funding allocation, the share every supplier may withdraw, or a recovery probability. The dated numerical companion is a reader-safe subset; unestablished values stay null rather than becoming zeros.

Read the evidence at its actual strength

Every promoted number needs a unit and a denominator

A share is not a dollar. A USDe-per-sUSDe conversion can be correct while the dollar price or recoverability of USDe differs. Aave’s measured valuation combines that ratio with a capped USDT/USD input. Its accepted value is not a liquidation sale quote. The relevant contracts, raw scales and formula are in the Aave case.

A concentration needs its denominator. The 88.761411% figure refers to the selected Base market allocation divided by one vault’s total assets at Base block 51,477,892, 18 September 2026 at 15:25:31 UTC. It does not measure USDe backing or the share of the ecosystem at risk. The source quantities and the separate market-supply-share ratio are in the same-batch table.

A report period is not a retrieval date. The $501 million total RWA amount is an issuer-reported June figure published on 17 July 2026. Reading it in September does not update it to September, and it is not all JAAA/STAC. Dated backing evidence.

Hypothetical means hypothetical. The normalized 1,015-assets/1,000-liabilities book and chosen stress haircuts are teaching examples from the research. They are not a reserve estimate or a forecast. The reserve is already included in the asset total; separate scenarios must not be stacked without changing their assumptions. Inputs and results.

Keep state, source and time attached

The investigation’s Ethereum and Base observations were provider-mediated reads. Some multi-call batches included their own block and timestamp getters. Those observations share a batch; the entire investigation does not share a single block or reporting time. On-chain timestamps are not the wall-clock times at which a researcher opened a document or received a response.

An ordinary explorer link is an identity locator, not a frozen replay of the displayed observation. To independently reproduce a past-block read, a reader needs a suitable archival data source and the same address, function, arguments and unit conversion. The present investigation’s calls were current-only. The foundation evidence trail states the binding and source limits rather than claiming archival replay.

A runtime hash proves identity of bytes, not what every function means. A common proxy-pattern screen cannot exclude every proxy design. A successful ABI call proves its returned result, not that an entire candidate source is equivalent. Conversely, a rejected or unavailable read does not supply a default value such as “no role,” “unpaused” or “no deployment.”

A token log, a balance and a recovery chain prove different things

The 20 September payment example matches a standard USDe burn to an outbound USDC Transfer in the same transaction and block, marked nonremoved. That is stronger than a token symbol or an unexamined Mint event. The investigation did not separately obtain the receipt/status; its candidate Mint-event signature did not match the observed topic. The retained candidate events are therefore not an ABI-verified redemption count or a throughput series. The precise pairing and its limits.

Chronology prevents double counting. That payment precedes the later Mint balance snapshot, so subtracting it again from the snapshot would duplicate the outflow. The burn and payout are two legs of one operation, not two independent cash payments. Neither log identifies a Maple or JAAA recovery as its source. Inventory versus payment evidence.

Denominators also prevent overstatement. Maple’s 88.64281% is the selected wallet’s share of that pool’s shares. JAAA’s 99.996563% is a share of the sampled Base token supply, not the whole fund. A JAAA token count is not a dollar NAV, and the pool’s direct USDC is not total withdrawable liquidity. Both observations are dated 20 September 2026 at separate blocks. Maple inputs · JAAA inputs.

A source label identifies why a wallet was investigated; it does not prove its beneficial owner or the legal agreement binding the position. The 14 September assessor values, published on 18 September, remain separate from 20 September token observations. No aggregate reserve ratio is reconstructed from inconsistent assessment denominators or an old NAV. Selection and evidence dates.

A useful cash test is conservative about identity, not automatically pessimistic about recovery. Track the paying entity, asset, network, release condition and deadline. A pool-to-issuer transfer moves one resource; it does not create a second reserve. A bridge adds cash and a debt or assigns proceeds. A positive final balance does not erase an earlier missed deadline; the supplied timing example checks the largest cumulative deficit. Cash conservation and maximum-prefix example.

A policy, an audit and a deployed observation answer different questions

The 22 September investigation reads public documentation, governance disclosures and published audits. Its later access date does not refresh the 18 and 20 September contract observations. A reported appointment remains distinct from agreement execution; a reported passed vote from activation; and conditional test approval from production completion. New documentary findings.

The Pashov Mint V2 Markdown report was read completely; Code4rena scope/summary, relevant dispositions and architectural discussion were inspected, not every line or embedded proof. In the 22 September documentary study, the original Guardian PSM audit was not obtained—only Kairos’s account. The later source-based continuation locates the auditor’s narrative and original artifacts, but does not inspect the PDF bodies. An audit-index entry is not an inspected report, and a fixes-review commit does not prove present deployment equivalence. Version identities and inspection limits.

The trust-summary conflict is preserved rather than averaged away: documentary seven-signature/seven-day statements are not the dated five-signature/one-day observations, nor is committee deliberation a contract delay. This is a source/version problem, not proof of compromise, a new configuration change or a universal immediate path. Disclosure conflict.

Partial is a scope statement, not an adverse safety verdict. The documentary work is usable, but current Mint/controller correspondence, Base transition semantics and remote controls remain incomplete. Missing private evidence does not show that the underlying agreement or safeguard does not exist. Decision-changing gaps.

An exact source identity does not establish deployment applicability

The 26 September comparison identifies the same MetaOracle Git blob at two differently linked repository commits. That supports a specific exclusion: changing between those two source pointers cannot explain the earlier recorded deviation mismatch. It does not establish that either revision was installed at the selected Base address. A factory’s initialization function, a factory locator and an instance’s actual creation history remain different evidence. Exact comparison.

Address-indexed source has a separate strength and limit. The provider returned the controller’s custom body and consequential inherited files, enabling direct reasoning about operation classes. The provider’s source/address attribution is not an independently rebuilt runtime or a present role roster. The full Mint target cannot be substituted by a controller, client ABI or an audit of another product. Controller-source scope.

Conditional reasoning must stay visibly conditional. The examined candidate’s view-only fallback does not persistently change selection; explicit challenge/healing operations have separate conditions. Those facts do not become installed F4 guarantees while correspondence remains unresolved. Likewise, successful controller call handling is not an economic receipt and does not resolve the earlier Mint-event/receipt gap. Policy matrix; call versus payment evidence.

Later publications can revise a status without proving completion: the 28 July reviewer statement advances the PSM record to reported fresh deployment, not verified production readiness. The newly located original audit PDFs remain unread; conflicting publisher summary counts are attributed, not combined. The complete 22 September chapter and older observation dates are preserved. Artifact discovery and product/version boundaries.

Match the cash sequence without turning balance changes into fees

The 7 October Aave result combines nine unique event records with adjacent historical end-of-block states and a later debt observation. Transaction hash, block hash, reserve/beneficiary fields and log order support the particular funding sequence. Protocol Borrow and Repay events corroborate actual transfers; counting them as extra movements would double-count cash. Recorded order and source trail.

USDC and USDT are reconciled separately. The difference between 10,006.329226 USDT paid and 10,001.640594 USDT received is the supported 4.688632-USDT own-cash contribution—not a fee. A 0.281595-USDT pre/post debt residual is an interval accrual/rounding distinction, not an inferred service charge. No full receipt or gas/sponsorship record supplies all-in expense.

The event at 05:44:35 UTC is not the later state at 14:04:11 UTC. The later account batch has inner block 26,140,985 and outer head 26,140,986; the inner block governs the observation. Adjacent end-of-block reads are not a full internal transaction trace, and incomplete intervening history does not attribute every subsequent debt change to interest. Event versus later debt.

The wider PT route uses exact successive integer quantities at the same explicit block, but sDAI previewRedeem is not a DAI receipt. Its 342,677.470045-USDC final quote remains an indicative alternative; the before-extra-cost gap of 557,322.529955 relates only to the hypothetical 900,000-USDC principal and that path. Old full-budget routes and the new one cannot be added together.

Finally, the selected shareholder’s successful 2,000-block query with no standard-withdrawal matches is not the same as the unavailable historical fee-denominator read. The former supports a bounded event negative; the latter supplies no value. Neither supports a failed-withdrawal or zero-expense claim. Scope and cost limits.

The new dated reader dataset contains selected recorded quantities and qualifications. Older paired files and sources remain unchanged; presentation checks do not independently validate financial evidence.

An upper envelope is neither a quote nor a fee invoice

5 October analysis uses preserved October 4 observations. A deliberately favorable relaxation can establish a useful conditional ceiling without optimizing a split or executing a sale. The new comparison grants all first-hop stock from two selected corridors, assumes no replenishment, ignores competition and extra costs, and sets at most one owed-token unit per intermediate token. Its assumptions travel with every promoted ceiling and gap. A smaller real opportunity cannot exceed that ceiling while those conditions hold; a wider resource set or different conversion ceiling is a different claim. Derivation and scope.

Do not add the PT and Aave envelopes: they are independent alternative allocations of the same stocks. Do not add downstream USDC without the intermediate input, or a free-token sale on top of the unchanged pool cash it would spend. A financed origination charge increases debt without increasing spendable proceeds; repayment of a replacement lender belongs in the complete ledger. Creditor sequence.

For costs, establish product identity, fee base, recipient and evidence type. The Coinbase lending help link names another Base vault. The described 1% spread concerns MORPHO rewards, not principal. The October 4 fee-share mint can be normalized at its own transaction’s withdrawal ratio, but the resulting approximately one-quarter-of-interest fraction is not a new 25% parameter, cash expense, gas measurement or selected-holder dilution. Source applicability and event inference.

The new source retrieval date does not refresh earlier source dates, normal admission or financial state. Missing account histories and receipts remain unknown, not zero. October 4’s natural other-holder payment already precedes the later snapshot; it is not a new deduction from that stock. No independent source or contract verification is implied by editorial presentation.

Track the cash already spent—and identify the obligation that ends

A balance, a quotation and the later use of its proceeds can describe one resource. The 3 October analysis using October 2 Aave inputs credits the selected pool’s 838,081.997038 USDT once. The 123,910.406881-USDT free-USDe quotation is a claim on that same stock, not an extra buffer. Subtracting it leaves 714,171.590157 USDT in the non-replenished single-pool scenario for later payment. Exact resource ledger.

Scope the conclusion to the sequence. An additional prior advance in a repay-all-first case is not the minimum financing required for every possible staged route. An independent equity-funded repayment can end the original loan without requiring a sale of all released collateral. A credit-funded repayment instead leaves the replacement creditor’s deadline and net proceeds unresolved.

Positive normal admission and uncertain execution costs are compatible. Missing cost evidence is not a failed withdrawal, and a conservative native maximum is not a cash reading. Conversely, a documented fee exclusion does not prove zero all-in expense. Keep the cost asset, payer, reimbursement and final received token distinct rather than inventing an ETH-to-USDC conversion. Shareholder amount and access boundaries.

Fixed-block quantity joins and natural payments answer different questions

The 4 October Ethereum comparison requests later financial views at the same block as the recorded starting quantities, 26,117,487. The inner clocks agree. A provider’s later outer node height is not the observation block and is not used to redate the result. Exact previous raw outputs size the next inputs; displayed decimal rounding is not fed back into the enquiry. Clocks and inspectable quantity trail.

This improves on a cross-block indicative path but does not execute a composite trade or reserve a future price. Staking conversion still needs a qualifying request and actual release. Prior funding that consumes a named pool’s cash changes the state available for later quotes. The full Aave comparison therefore quotes the combined collateral-conversion and free-USDe amount once, rather than adding two outputs from the same untouched inventory. The non-additive account comparison.

A natural withdrawal is a different evidence class. Token transfer, vault withdrawal and share burn can be matched by block/transaction hash, amount, owner and recipient. A separate fee-share mint may be decoded through its published event interface without becoming a cash charge or proving complete implementation correspondence. Here the observed payment belongs to another holder, and no independently retrieved receipt or gas/sponsorship record establishes selected-holder all-in expense. Payment and claim creation.

Chronology is part of cash conservation: the matched payout predates the later Base snapshot, so deducting it again double-counts the use. A stored market timestamp can likewise precede its retrieval: the new Base debt uses accrual recorded 32 seconds earlier and is not a future payoff invoice. Earlier chapters keep their original input clocks. Stored debt and ownership boundary.

A correct quantity join is necessary—but does not make different blocks one executable route

The 2 October investigation establishes an actual Aave direct-sUSDe/USDT account and a particular shareholder’s shares and normal admission. Those are new financial readings. They do not retrospectively refresh the September USDe-debt accounts, the older eight-market list or the 1 October analysis that deliberately used frozen inputs. Clocks, identities and source scope.

The PT comparison makes the distinction visible. At Ethereum block 26,103,799, the USDe input produced 837,882.396443 USDT. The later second-leg enquiry at 26,103,809 uses exactly that amount and returns 837,486.736607 USDC. An earlier first-leg quote returned 916,509.044895 USDT; its second-leg result cannot be used as the proceeds of the later, smaller output. Both original enquiries are retained rather than averaged or silently selecting the favorable one.

The matched result is still a cross-block indicative path. Native PT output, asset-conversion value and the two cash legs have separate clocks; a staking wait can separate an actual holder from the USDe being quoted. Exact unit arithmetic does not reserve liquidity, establish consumed input in a boundary-limited quote or fund the interval. Correct and excluded quantity joins.

Two other distinctions now have concrete examples. Outside-Aave inventory can supply principal without releasing pledged shares first, but a token balance does not prove unrestricted use, native-cost funding or actual repayment. Normal receiver admission can return true while bundle allowance is zero and execution remains unfunded. A zero conservative maxWithdraw is not a measured zero cash balance. Principal before costs · Admission, allowance and cash.

The date of an analysis is not the age of its inputs

The 1 October financing continuation adds public financial-route documentation and calculations without new on-chain readings. Its Base balances, borrower debt, PT quote and vault screens remain September 28 inputs, often from different batches. Freezing them for a clearly labelled comparison is not reconstructing a simultaneous historical snapshot.

A conservative conclusion can use optimistic assumptions, but the scope must remain visible. The Base model gives one borrower all cash in a finite four-pool set and assumes zero costs. It then asks whether that set could settle the recorded debt without additional net cash. The answer does not value leftover collateral at zero, exclude all other outlets or establish an actual shortfall today.

Separate three amount questions: an offered trade budget versus input actually consumed; a native output or accounting equivalent versus the creditor’s token; and gross share-value debit versus net receipt. A share penalty retained in a vault is not outward USDC, while an external fee is. A gross/net approximation is neither an SDK return nor an executable withdrawal maximum.

Admission also requires the correct object. A published route address is not a measured account, allowance is not a share balance, and old sender-side gate readings did not establish the bundle’s receiver admission. The 2 October receiver checks now answer that normal-view question for the sampled configuration; authorization, costs and future access remain separate. A same-resource repayment is not new independent funding. The additional calculation checks those financial distinctions; it does not test contract behavior or create the missing financing evidence. Receiver boundary · Amount conventions · Cash conservation.

Read what the return actually measures

The examples in this section retain the 28 September evidence boundary. The newer quantity-matched PT path and actual USDT account above advance specific propositions without rewriting these dated examples.

Input budget is not automatically consumed input. The 28 September million-unit Base quotation returns its default extreme price boundary. Its output fields do not report consumed input, so no full million-unit sale price or realized loss is established. Unknown leftover is neither written off nor counted as cash. The failed full-position quote supplies no output, not zero proceeds. Exact returned fields and interpretation.

Account denomination is evidence, not a label to substitute. The two actual Aave accounts owe USDe and use eMode 32. They do not turn September 27’s hypothetical USDC/eMode-2 account into a measured one. The relevant reserve flags and debt balances were reconciled; a category name mentioning a PT maturity is not its complete eligibility list. Measured debt and collateral.

An accounting conversion is not another quote. The PT route returns net sUSDe. Multiplying it by the same-batch SY ratio gives a USDe accounting equivalent, not USDC or an independently quoted next leg. The reported SY fee is not deducted again from net output. Native output and remaining payment.

A reconciled portfolio still needs an access test. The eight expected claims sum to returned vault assets. Capping each market’s nominal unborrowed balance by the vault’s claim prevents one overcount, but it does not reserve cash for a shareholder or bypass the selected withdrawal route. The documented share penalty is not outward USDC cash. The screen and its actual boundary.

The evidence’s clock is 28 September UTC, even where the user’s local calendar has reached September 29. Included block timestamps are not retrieval times, and a later repayment used in a conditional ledger is not same-state inventory. Older chapter dates and downloads do not change when this source interpretation improves. Observation clocks and retained downloads.

A financial bound, a quote and a funded payment are different evidence

The 27 September investigation uses ordinary financial observations, arithmetic on those inputs and explicitly hypothetical schedules. The observed Base borrower is not the hypothetical Aave account. A no-repayment release formula describes a necessary fixed-input financial bound, not a measured executable withdrawal limit. Contract rounding, elapsed interest, admission and transaction inclusion remain outside that formula. The observed inputs and bound.

The Ethereum pool’s four size quotations are independent price views. They do not consume inventory, establish a minimum received amount, or add up to sequential exit capacity. The pool’s own balance getters are coin-accounting records, not independently queried ERC-20 balances. Freezing a September 27 quote as a sale after cooldown is an assumption that must remain labelled. Route and observation boundary.

The term-funded unwind makes the initial lender, amount, asset and timing explicit. It is still assumed funding: no real facility or complete executed chain was established. Where sale proceeds fall short, the balancing ledger requires an additional contribution; a negative residual cannot be described as full payment. A loan repayment converts an existing receivable to cash and is counted once. Creditor transfer and cash conservation.

Freshness also belongs to the returned record. The active PT market was read on 27 September but returned a stored accounting update from 25 September, 20:03:47 UTC. Separate batches, price paths and networks are not a common balance-sheet instant. The one-SY preview is not an executed future PT redemption. These distinctions survive even when all arithmetic is internally consistent. Evidence clocks and selected observation download.

A mismatch is a finding about the evidence

The Base meta-oracle provides a concrete example. The inspected candidate source’s deviation formula does not reproduce the observed getDeviation() result using the recorded primary and backup. The public assessment therefore preserves the measured inputs, threshold and timers, but withholds a complete claim about fallback and healing semantics. Another formula matching one result would not prove all of the implementation.

This is neither a demonstrated exploit nor a reason to pretend the candidate is a complete deployment match. It identifies the evidence needed for a stronger conclusion. The numerical comparison remains visible in the main oracle discussion, not hidden in a technical appendix.

Likewise, an unread attestation is not an adverse opinion. A proposal is not a deployed safeguard. A reported historical redemption is useful counterevidence, but not an independent transaction replay. Those distinctions prevent uncertainty from being turned into either automatic reassurance or an unsupported failure claim.

Corrections belong to propositions and dates

Dated observations do not change when a reader opens a page later. The control report dated 12 September uses an 11 September snapshot; the 18 September whole-system assessment separates its 17–18 September state observations from older portfolio reporting. Neither date should be silently assigned to the other’s evidence.

Differences in inspected source material are not themselves evidence that a deployment changed or that an earlier conclusion was false. The useful correction identifies the old proposition, the added evidence and the remaining limit. All dated reports remain inspectable, and History preserves their surrounding context.

The reader pages require no wallet connection, keys, transaction or background data request. Source links are deliberate references to open; the tables and scenarios are static.