Evaluated on: · Spec revision: 2026-10-07 · 6dac5e7491 · EIP-8081 list: SFI
Scope at the cutoff. EIP-8141 adds an EIP-2718 transaction type (0x06), the frame transaction. Its body is an explicit sender, a list of frames (DEFAULT, VERIFY or SENDER mode, with flags, target, per-frame execution and state gas limits, value and data), a list of protocol-validated signatures (ARBITRARY, SECP256K1, P256) and optional blobs. Validity and payment are decided by EVM execution: VERIFY frames call a new APPROVE opcode to set sender approval and the payer. Any VERIFY failure, or a payer that is still unset after all frames, makes the transaction invalid. It defines per-frame two-dimensional gas pools with cross-frame state-gas refill attribution, a new intrinsic and calldata-floor settlement, block reservations per dimension, a new per-frame receipt format, atomic batches, default code for code-less accounts, an expiry verifier contract, and six introspection opcodes. It also changes ORIGIN semantics and exempts frame transactions from EIP-3607, and it specifies public-mempool and networking rules.
- Evaluator
- LLMChecklist v3
- Confidence
- Medium
- Under-specified at assessment cutoff
- Yes — 5 criteria affected
- Plausible range
- 50–57 (High)
- Snapshot
- 2026-10-07 · EIP revision
6dac5e7491(2026-10-07)
Score bands · Checklist revision 3
- Low <12
- Medium 12–22
- High ≥23
28 criteria scored 0–3 (4 in exceptional cases; cross-EIP interactions is uncapped); nominal maximum 84.
Complexity profile
Each segment is one criterion's contribution to the LLM total. Hover or focus a segment for its score and rationale.
Top complexity drivers
- Cross-EIP interactions5
- Added opcodes4
- New or modified transaction validity mechanisms4
- Modified opcodes3
Under-specified at assessment cutoff: Yes
The EIP text available at the assessment cutoff left material behavior unresolved. The affected criteria and the plausible total range record that uncertainty.
Why: Several localized outcomes are not fully specified: whether payer and sender_approved are rolled back when APPROVE runs in a sub-call that later reverts inside a successful frame; how the frame-entry access charge is ordered against the SENDER approval-validity check; effective_gas_price and coinbase payment for frame transactions; and how frame-transaction blob hashes relate to the Engine API.
Plausible total
50–57
recorded score 52 · plausible tiers High
Unresolved questions at the cutoff (4)
- Is the approval context (payer, sender_approved) journaled per EVM call frame, so that it reverts when an enclosing sub-call reverts?
- For a SENDER frame with sender_approved == false whose limits.execution cannot cover the entry access charge, is the result an exceptional halt or an invalid transaction?
- How are effective_gas_price and the coinbase priority fee defined for frame transactions?
- Must the Engine API's expected blob versioned hashes include those carried by frame transactions?
Notable ambiguities noted by the assessor (5)
- APPROVE is reachable at depth > 0 (self-call or DELEGATECALL keeps ADDRESS == resolved_target), but approval-context journaling is defined only at the frame level.
- "Being a frame target does not warm an address" vs "journaled like any other EIP-2929 access": the warm status of a target after the frame succeeds is open to more than one reading.
- Interaction of the frame-entry charge with SENDER approval validity: the ordering lists the charge first.
- A deploy frame setting the sender's nonce (EIP-161 contract nonce = 1) followed by an APPROVE increment makes the next expected nonce non-obvious.
- Clients may skip EVM execution for the expiry verifier frame, but gas consumed must match canonical execution.
Criterion breakdown
| Criterion | Score | Why this score | Evidence / uncertainty |
|---|---|---|---|
| Cross-EIP interactionsExceptional | 5 | The target couples multiple EIPs' behavior inside one settlement and frame-execution model. That requires coordinated scenarios across them, so level 3. |
Confidence: High Interacting EIPs: EIP-8037, EIP-2780, EIP-7976, EIP-7778, EIP-3529, EIP-4844, EIP-7702, EIP-7708, EIP-2929, EIP-1559, EIP-2718, EIP-3607, EIP-3651, EIP-6780, EIP-7594, EIP-7825, EIP-8038, EIP-2, EIP-7819, EIP-7997 |
| Added opcodesExceptional | 4 | Seven instructions are added, several of them complex. The level-3 condition is met. Exceptional score: Level 3 needs only two instructions with one complex. This adds seven. Two of them (TXPARAM, FRAMEPARAM) have 13 and 12 selector values each, with per-selector halt rules (current/future frame, scheme-dependent halts). SIGDATACOPY and SIGPARAM have scheme-conditional availability. APPROVE carries a full approval state machine with balance, nonce and state-gas effects. All seven halt outside frame transactions. This is opcode testing work well beyond the level-3 threshold. |
Confidence: Medium Uncertainty: Using the exceptional level is a judgment call. |
| New or modified transaction validity mechanismsExceptional | 4 | Validity is now decided partly by EVM execution across frames. Tests need restructured validation sequencing and coordinated state scenarios (e.g., a DEFAULT frame changing state that a later VERIFY frame depends on). Exceptional score: Beyond the level-3 restructuring, validity spans four distinct layers that must all be tested: static structure, cryptographic signature validity, execution-dependent outcomes after arbitrary frames (any VERIFY failure anywhere, a SENDER frame before approval, payer unset at end), and per-dimension block capacity. A block becomes invalid because of a VERIFY or APPROVE outcome reached only after executing earlier state-modifying frames. Each layer needs its own coordinated transaction/state scenarios. |
Confidence: Medium Uncertainty: Using the exceptional level is a judgment call. |
| Modified opcodes | 3 | Existing instruction semantics change: ORIGIN return values and transient-storage lifetime. |
Confidence: High |
| State gas accounting changes | 3 | Introduces a new state-gas charging and attribution mechanism (frame-scoped pools, owner-frame refills, receipt mutation with rollback) and new charge sites, replacing the baseline reservoir/spill accounting. This meets level 3. |
Confidence: High |
| New transaction types | 3 | Introduces a distinct transaction envelope. |
Confidence: High |
| Encoding changes (RLP/SSZ) | 3 | New transaction, receipt and p2p schemas. |
Confidence: High |
| Block syncing changes | 3 | Block import must decode a new transaction type and apply several structural rules, including complex cross-transaction reservation checks. |
Confidence: Medium Uncertainty: Some of these checks could be classed as ordinary execution rules, which would lower this to 2. |
| Security risks | 3 | Shared authorization and trust invariants change across components: the transaction pool, EVM approval, payer settlement, ORIGIN-based checks and sender identity. This needs coordinated adversarial scenarios. |
Confidence: High |
| Performance risks | 3 | Several new workload couplings need integrated stress testing across subsystems: signature verification per entry (P256/secp256k1), up to 64 frames with per-frame receipts hashed into the trie, execution-dependent invalidity during block building, and mempool simulation and revalidation. |
Confidence: Medium Uncertainty: No benchmark data was supplied. |
| Edge/boundary conditions | 3 | Many boundary-sensitive rules are added. APPROVE outcomes and frame dispatch form an elevated matrix: mode × flags/scope × target resolution (null/sender/other, code/delegated/empty/precompile) × approval state × balance and pool exhaustion. |
Confidence: High |
| EVM Gas rule changesUnder-specified | 2 | Adds a new execution-gas accounting mechanism for the new envelope: per-frame pools, the frame-entry access charge, intrinsic decomposition, signature-verification gas, floor/settlement and per-dimension block reservation. Gas results for baseline operations in existing transaction types are unchanged, so level 2. |
Confidence: Medium Uncertainty: Opcodes in frame transactions run under different warm-address initialization and settlement (no tx.to warming, ORIGIN cold). This could be read as changing existing accounting rules (level 3). |
| State-access ordering within opcode execution | 2 | New state-accessing operations (frame dispatch, APPROVE) need their own ordering rules, and there is a BAL exclusion for protocol signature checks. No existing opcode class's ordering changes, so level 2. |
Confidence: Medium Uncertainty: EIP-7928 is not supplied, so the full BAL-recording consequences of frame-entry accesses cannot be verified. |
| Transition-tool interface changesUnder-specified | 2 | Several semantic input and output fields change (transaction structure, signatures, payer, per-frame receipts). There is no clear new exchange mechanism, so level 2. |
Confidence: Medium Uncertainty: No transition-tool evidence was supplied. Reporting execution-dependent rejections might require a new mechanism (level 3). |
| New test-framework primitives | 2 | Requires new construction abstractions (frame transaction builder, multi-scheme signer, per-frame receipt and two-dimensional gas expectations) within the target's suite. |
Confidence: Medium Uncertainty: If frame transactions must become a shared transaction-type parameter for existing opcode/state-gas families, this reaches level 3. |
| Cryptography | 2 | Several established signature mechanisms (secp256k1 and P256 transaction signatures, with new canonicality and address-derivation rules) are added to EL transaction validation. |
Confidence: Medium Uncertainty: Test resources for P256 are only indirectly evidenced (EIP-7951 is not supplied). |
| Unspecified behavior requiring cross-client consensusUnder-specified | 2 | Several localized points have competing plausible outcomes that need spec or client agreement. |
Confidence: Medium Uncertainty: Supplied dependency text might resolve the fee details, for example by analogy with EIP-1559. |
| Added system contracts | 1 | Introduces exactly one stateless protocol-designated contract with no system action. |
Confidence: Medium Uncertainty: It could be treated as an ordinary contract because it is deployed normally, which would make this 0. |
| Blob gas accounting changesUnder-specified | 1 | Existing EIP-4844 blob accounting is reused. Only the funding/settlement parameters change (the payer rather than the sender, and max_cost computed at blob_base_fee). No new blob mechanism is added. |
Confidence: Medium Uncertainty: The change in payer and balance-check basis could be read either as a validity-only change (0) or as a new settlement path (2). |
| Patterns affecting pre-existing tests | 1 | Baseline behavior of existing transaction types is preserved. Rework is limited to opcode-definition boundary cases (newly allocated byte values listed as valid at the fork) within one family. |
Confidence: Medium Uncertainty: Behavior without a test suite is estimated, not measured. |
Show 8 zero-score criteria
| Criterion | Score | Why this score | Evidence / uncertainty |
|---|---|---|---|
| Added precompiles | 0 | No new precompile. |
|
| Modified precompiles | 0 | No precompile semantics or gas change. |
|
| Modified system contracts | 0 | No existing system contract changes. |
|
| New EVM gas refund | 0 | No new refund-counter mechanism is introduced. The cross-frame state-gas refill attribution is scored under state gas accounting. |
Uncertainty: The cross-frame refill crediting via owner receipts is history-dependent. It could be classified as a complex new refund (level 2). |
| New block / header fields | 0 | No new header field. |
|
| New fork activation mechanism | 0 | No activation-specific state transition. |
|
| Engine API changesUnder-specified | 0 | Transactions remain opaque in payloads, and no Engine API field or endpoint is specified. |
Uncertainty: The EIP does not say how frame-transaction blob versioned hashes interact with newPayload's blob hash checks. |
| New invariant on pre-existing tests | 0 | Baseline tests on existing transaction types gain no new output to assert. |
|
Assessment provenance
- Assessed EIP revision
ethereum/EIPs@6dac5e7491EIPS/eip-8141.md committed 2026-10-07 · information cutoff 2026-10-07T22:23:55Z- Rubric
- Checklist revision 3 ·
ethspecs/pm@fe2f793b03 - Evaluator
- Opus 5.5 (
claude-opus-5-5) at high effort, one tool-less call per EIP · isolationbubblewrap_claude_p_no_tools_v1 - Source record
- Frozen research record
research/tasks/10-opus-v3-reassessment/prospective/outputs/assessments/hegota-2026-10-08/eip-8141.yaml· sha256de6dfa2e1879 - Supporting documents supplied with the EIP
supporting/eip-20.md,supporting/eip-1559.md,supporting/eip-2718.md,supporting/eip-2780.md,supporting/eip-2929.md,supporting/eip-3529.md,supporting/eip-3607.md,supporting/eip-3651.md,supporting/eip-4337.md,supporting/eip-4844.md,supporting/eip-6780.md,supporting/eip-7594.md,supporting/eip-7702.md,supporting/eip-7708.md,supporting/eip-7778.md,supporting/eip-7819.md,supporting/eip-7825.md,supporting/eip-7843.md,supporting/eip-7976.md,supporting/eip-7997.md,supporting/eip-8037.md,supporting/eip-8038.md,supporting/erc-7562.md