Retrospective LLM-Based Complexity Evaluations

Prospective · frozen 2026-10-08 snapshot

Hegotá candidate complexity assessment

EL-rubric total by EIP-8081 list: Scheduled 77 (2 EIPs); with Considered 385 (19); with Proposed 513 across 25 scored EIPs, using the lists after ACDE 247. 8 entries are not applicable to the Execution Layer rubric. EIP-7666 and EIP-8355 were declined (DFI) and remain listed below, excluded from every total.

Independent LLM Evaluation. These scores were produced with Claude Opus 5.5 under checklist revision 3, the same LLM and rubric as the retrospective study, independently of the STEEL team’s ongoing manual assessment work. The Human column below records where that manual work stands for each candidate.
EIP-8081 lists as of ACDE 247 (EIPs f154af8, 2026-10-08). Scores are unchanged from the assessment at 6dac5e7. Not yet assessed, so not shown or counted: EIP-7907 Meter Contract Code Size (PFI), EIP-8360 TCREATE Opcode (CFI).
Common information cutoff: 2026-10-07T22:23:55Z. The assessment covers Execution Layer and execution-client networking surfaces only. It does not predict final fork scope or include observed outcomes. Combine the lists into a scope in the Hegotá scope builder.
Scored EL surfaces
27
8 not applicable
Low / Medium / High
6 / 12 / 9
Under-specified at snapshot
26 / 27
Human checklists
15 of 35
  • Complete 2
  • Available in open PR 8
  • Draft PR 5
  • Pending 20
15 Human/LLM comparisons
Human checklist sources: ethspecs/pm at 2026-08-14 plus 24 open pull requests (captured 2026-09-13)

Statuses follow the real source state. Complete is merged on main; Available in open PR exists only in an open pull request; Draft PR is in a draft pull request; Incomplete has unresolved cells or inconsistent totals and carries no score; Pending has no checklist anywhere. A missing checklist is never a zero.

Where the Complexity Comes From

Each segment sums one criterion across the fork's scored LLM assessments. Hover or focus a segment for its points, share, and the number of contributing EIPs.

View
Hegotá513 points · 25 EIPs
Composition table: points per criterion and fork
Hegotá: criterion points across 27 scored candidates
CriterionHegotá
Added opcodes13 · 3% · 5 EIPs
Modified opcodes15 · 3% · 5 EIPs
Modified precompiles3 · 1% · 1 EIP
Added system contracts9 · 2% · 5 EIPs
Modified system contracts4 · 1% · 3 EIPs
EVM Gas rule changes18 · 4% · 12 EIPs
State-access ordering within opcode execution15 · 3% · 8 EIPs
Blob gas accounting changes5 · 1% · 3 EIPs
State gas accounting changes14 · 3% · 7 EIPs
New EVM gas refund2 · 0% · 1 EIP
New transaction types3 · 1% · 1 EIP
New or modified transaction validity mechanisms20 · 4% · 10 EIPs
New block / header fields3 · 1% · 1 EIP
Encoding changes (RLP/SSZ)33 · 6% · 11 EIPs
Block syncing changes18 · 4% · 9 EIPs
New fork activation mechanism9 · 2% · 3 EIPs
Engine API changes9 · 2% · 5 EIPs
Transition-tool interface changes16 · 3% · 11 EIPs
Patterns affecting pre-existing tests40 · 8% · 23 EIPs
New invariant on pre-existing tests10 · 2% · 6 EIPs
New test-framework primitives30 · 6% · 21 EIPs
Security risks40 · 8% · 24 EIPs
Performance risks29 · 6% · 19 EIPs
Edge/boundary conditions51 · 10% · 23 EIPs
Cryptography6 · 1% · 3 EIPs
Cross-EIP interactions59 · 12% · 23 EIPs
Unspecified behavior requiring cross-client consensus39 · 8% · 24 EIPs
Total513

Complexity Assessments

LLM rows were produced by Claude Opus 5.5 at high effort, evaluating the EIP revision linked in the Assessment Input column against revision 3 of STEEL’s complexity-assessment template. Human rows are STEEL reviewers’ published checklists; their status badge links to the pull request or merged report. Select rows from either evaluator to compare them, or open an EIP for its full assessment.

All EIPs were evaluated at the same ethereum/EIPs commit, 6dac5e7. Human checklist evaluation dates are not recorded.

Evaluator
42 assessments: 27 LLM, 15 Human· 20 Hegotá candidates have a pending human checklist
Hegotá candidates at EIPs 6dac5e7: one row per evaluator, with not-applicable entries and pending human checklists counted above
Select for comparisonComplexity profile
EIP-3298 Remove storage-clear refund and refund capCFILLMChecklist v3Complete10LowNo
EIP-3298 Remove storage-clear refund and refund capCFINot recordedHumanChecklist v2Open PR6Low—
EIP-4758 Deactivate SELFDESTRUCTCFILLMChecklist v3Complete11LowYes
EIP-4758 Deactivate SELFDESTRUCTCFINot recordedHumanChecklist v2Open PR8Low—
EIP-5920 PAY opcodeCFILLMChecklist v3Complete18MediumYes
EIP-5920 PAY opcodeCFINot recordedHumanChecklist v2Open PR14Medium—
EIP-7666 EVM-ify the identity precompileDFILLMChecklist v3Complete18MediumYes
EIP-7666 EVM-ify the identity precompileDFINot recordedHumanChecklist v2Draft PR12Medium—
EIP-7668 Remove bloom filtersCFILLMChecklist v3Complete12MediumYes
EIP-7668 Remove bloom filtersCFINot recordedHumanChecklist v2Open PR7Low—
EIP-7709 Read BLOCKHASH from Storage and Update CostCFILLMChecklist v3Complete14MediumYes
EIP-7709 Read BLOCKHASH from Storage and Update CostCFINot recordedHumanChecklist v2Open PR17Medium—
EIP-7805 Fork-choice enforced Inclusion Lists (FOCIL)SFILLMChecklist v3Complete25HighYes
EIP-7805 Fork-choice enforced Inclusion Lists (FOCIL)SFINot recordedHumanChecklist v1Complete15Medium—
EIP-7906 Transaction Assertions via State Diff OpcodeCFILLMChecklist v3Complete27HighYes
EIP-7906 Transaction Assertions via State Diff OpcodeCFINot recordedHumanChecklist v2Open PR23High—
EIP-7979 Call and Return Opcodes for the EVMCFILLMChecklist v3Complete13MediumYes
EIP-8025 Optional Execution ProofsPFILLMChecklist v3Complete25HighYes
EIP-8077 eth/XX - announce transactions with nonceCFILLMChecklist v3Complete8LowYes
EIP-8116 Replace cumulative receipt fieldsCFILLMChecklist v3Complete9LowYes
EIP-8131 Unified Transaction Content FloorCFILLMChecklist v3Complete16MediumYes
EIP-8141 Frame TransactionSFILLMChecklist v3Complete52HighYes
EIP-8141 Frame TransactionSFINot recordedHumanChecklist v1Complete38High—
EIP-8142 Block-in-Blobs (BiB)PFILLMChecklist v3Complete41HighYes
EIP-8148 Custom sweep threshold for validatorsPFILLMChecklist v3Complete19MediumYes
EIP-8151 Account Code Restricted ecRecoverCFILLMChecklist v3Complete18MediumYes
EIP-8198 Quick SlotsCFILLMChecklist v3Complete17MediumYes
EIP-8205 Withdrawal credentials preregistrationPFILLMChecklist v3Complete16MediumYes
EIP-8205 Withdrawal credentials preregistrationPFINot recordedHumanChecklist v1Open PR5Low—
EIP-8250 Keyed Nonces for Frame TransactionsCFILLMChecklist v3Complete42HighYes
EIP-8250 Keyed Nonces for Frame TransactionsCFINot recordedHumanChecklist v2Draft PR22
published 20
Medium—
EIP-8253 Bump nonce of zero-nonce storage accountsCFILLMChecklist v3Complete11LowYes
EIP-8272 Recent Roots for Frame TransactionsCFILLMChecklist v3Complete24HighYes
EIP-8272 Recent Roots for Frame TransactionsCFINot recordedHumanChecklist v2Open PR14Medium—
EIP-8279 Block Access List Byte FloorCFILLMChecklist v3Complete24HighYes
EIP-8279 Block Access List Byte FloorCFINot recordedHumanChecklist v2Draft PR22Medium—
EIP-8298 SETCODEFROM Code Reuse InstructionCFILLMChecklist v3Complete34HighYes
EIP-8355 Precompiles for ML-DSA VerificationDFILLMChecklist v3Complete12MediumYes
EIP-8368 CPSB Recalibration for New Gas LimitPFILLMChecklist v3Complete11LowYes
EIP-8368 CPSB Recalibration for New Gas LimitPFINot recordedHumanChecklist v2Draft PR16Medium—
EIP-8372 Normalized state gas limitPFILLMChecklist v3Complete16MediumYes
EIP-8372 Normalized state gas limitPFINot recordedHumanChecklist v2Draft PR20Medium—

Scored proposals

Every scored assessment as a stacked complexity bar, ranked by total on one shared scale. Switch evaluator to rank the human checklists, or show both per EIP. The 8 not-applicable entries remain in the table above.

Evaluator
  1. EIP-8141 Frame TransactionSFIChecklist revision 3
  2. EIP-8142 Block-in-Blobs (BiB)PFIChecklist revision 3
  3. EIP-8025 Optional Execution ProofsPFIChecklist revision 3
  4. EIP-5920 PAY opcodeCFIChecklist revision 3
  5. EIP-8198 Quick SlotsCFIChecklist revision 3
  6. EIP-7668 Remove bloom filtersCFIChecklist revision 3
  7. EIP-4758 Deactivate SELFDESTRUCTCFIChecklist revision 3

Caveats

Download all assessmentsPublication JSON
Criterion legend and glossary

Every stacked bar, comparison matrix, and criterion table on this site uses the same criterion colours, abbreviations, and order. Colour marks the criterion group; the abbreviation and name identify the criterion. Scores are 0–3 per criterion (4 is exceptional; cross-EIP interactions is uncapped).

EVM surface

Opcodes, precompiles, and system contracts that are added or modified.

  • Added opcodes
    Introduces new opcodes
  • Modified opcodes
    Modifies pre-existing opcodes
  • Added precompiles
    Introduces new precompiles
  • Modified precompiles
    Modifies pre-existing precompiles logic or gas-accounting
  • Added system contracts
    Introduces new system contract, stateful or not
  • Modified system contracts
    Modifies pre-existing system contracts

Gas and accounting

Execution, blob, and state gas rules, refunds, and where charges happen inside opcodes.

  • EVM Gas rule changes
    New EVM gas accounting rules
  • State-access ordering within opcode execution · not in checklist revision 1
    Changes *where inside an opcode's execution* state is accessed, or where gas is charged relative to that access. Because a state access is recorded in the block-level access list only if execution had enough gas to reach it, this ordering is consensus-critical: moving it changes the BAL at every gas boundary of every affected opcode.
  • Blob gas accounting changes
    New Blob gas accounting rules which potentially affect pre-existing tests
  • State gas accounting changes · not in checklist revision 1
    New state gas accounting rules. State gas is the cost of *writing* state, as opposed to accessing or executing it: `StateGasCosts`, `COST_PER_STATE_BYTE`, the block-level state gas budget, and the spill path into execution gas.
  • New EVM gas refund
    New gas-refund mechanism

Blocks, transactions, and encoding

Transaction types and validity, block and header fields, encodings, syncing, and activation-time changes.

  • New transaction types
    Introduces a new transaction type
  • New or modified transaction validity mechanisms
    Creates new or modifies pre-existing transaction types' validation mechanisms
  • New block / header fields
    Introduces new block or block header fields
  • Encoding changes (RLP/SSZ)
    Introduces encoding changes at the transaction/block/interfaces level
  • Block syncing changes
    Modifies block RLP validation mechanisms that require test client syncing.
  • New fork activation mechanism
    Modifies state, internal variables, or similar, at the fork activation block

Client interfaces

Engine API and transition-tool interface changes.

  • Engine API changes
    Introduces new fields to the Engine API directives
  • Engine API encoding changes · Checklist revision 1 only
    Engine API encoding changes (the revision-1 template defines no anchor text for this row).
  • Transition-tool interface changes
    Modifies or adds new fields to the transition tool interface.

Testing impact

Rework, new invariants, and new primitives required in the test framework.

  • Patterns affecting pre-existing tests
    Implements a new validation mechanism or rule that translates in reworking pre-existing tests
  • New invariant on pre-existing tests · not in checklist revision 1
    Tests that are **not about this EIP** must nonetheless assert something this EIP produces. Their logic does not change; they gain a new thing to check.
  • New test-framework primitives · not in checklist revision 1
    Requires new abstractions in the test framework itself — expectation types, modifiers, helpers — beyond writing test functions with what already exists.

Risk and validation

Security, performance, boundary conditions, and cryptography that need validation.

  • Security risks
    Introduces or modifies mechanisms that could compromise the security of the chain, users, validators, or other stakeholders, if not implemented properly.
  • Performance risks
    Introduces or modifies mechanisms and requires performance validation.
  • Edge/boundary conditions
    Feature contains edge/boundary conditions.
  • Cryptography
    Introduces new cryptography mechanisms or modifies existing functionality that involves cryptography

Coordination

Cross-EIP interactions and behavior that clients must agree on before tests exist.

  • Cross-EIP interactions
    Introduces or modifies mechanisms that affect other EIPs in either the same or past forks.
  • Unspecified behavior requiring cross-client consensus · not in checklist revision 1
    The EIP text does not determine the answer for cases a test can construct. Clients must agree on a previously unspecified detail before tests can be baselined. The cost here is coordination and re-baselining, not test writing.