A concise reconstruction has not been recorded.
Inspect the ClaimsCase · DiggingBeagle record
EVAge autonomously generated and adapted MEV strategies across EVM chains
Research demonstration of EVAge, a hierarchical multi-agent harness that generates, repairs and adapts MEV-bot code offline, then validates candidate strategies deterministically on forked historical EVM state. The authors evaluate more than 1.5 million historical blocks from each of Ethereum, Base and BNB Smart Chain, report five HFT-composed MEV variants, cross-protocol adaptation to Uniswap V3 and Balancer V2, and cross-chain migration to Base and BSC. This is a controlled research benchmark, not a confirmed live attack: experiments did not broadcast transactions or report production losses, and the reported revenue figures are upper-bound replay estimates under strong capital and competition assumptions.
Evidence boundary
An explicit boundary statement has not yet been recorded. The evidence ledger defines what can be supported.
30-second account
No separate implication has been established in the canonical account.
Read the stated implicationsNo explicit unknowns are recorded. That is not evidence of completeness.
Inspect limits and uncertaintyFull canonical reconstruction
Mechanism and trust boundary
Typed chronology
Dates retain their recorded precision. Partially dated events can overlap; display order does not establish a causal sequence.
- 2025-02 (month precision) to 2025-08 (month precision)Event type unspecified
Historical EVM evaluation window
The authors report datasets of more than 1.5 million blocks for each of Ethereum, Base and BNB Smart Chain spanning February through August 2025.
- Sep 23, 2026Event type unspecified
EVAge preprint publicly disclosed
The EVAge preprint was submitted to arXiv and publicly disclosed.
- Sep 23, 2026disclosure
Public disclosure
Claims & evidence
14 independently addressable Claims. Expand a Claim to inspect support, contradiction and scope.
CLM-EVAGE-ABLATIONIn a 10,000-block BSC ablation using HFT, SBA and JIT, EVAge's specialized multi-agent harness found 4,546 profitable cases totaling 31.17 WBNB versus 109 profitable cases totaling 3.81 WBNB for the paper's pipeline-structured single-agent baseline; the authors attribute much of the gap to better candidate collection, BSC-specific planning and diagnostic role separation.supported
Basis: reported finding
- supportsEVAge: Autonomous MEV Generation and Adaptation via Multi-Agent Harnessprimary disclosure
Section 5.5 'Ablation Study'; Table 5 and accompanying findings
CLM-EVAGE-AUTONOMOUSWithin an EVAge run, an orchestrator coordinates collector, builder and validator-engineer subagents that iteratively generate and repair MEV-bot code without human intervention in the orchestration loop. The paper nevertheless states that reproduced reference-strategy code was further reviewed/censored by human experts, so the authors' 'fully autonomous' characterization should be read as applying to the configured harness run rather than every stage of research preparation.supported
Basis: reported finding
- supportsEVAge: Autonomous MEV Generation and Adaptation via Multi-Agent Harnessprimary disclosure
Sections 4.2-4.3, especially the orchestrator/subagent workflow; Section 5.1 'Reference strategies', paragraph stating reproduced code was further censored by human experts
CLM-EVAGE-EVALUATIONThe authors evaluate EVAge on more than 1.5 million historical blocks from each of Ethereum, Base and BNB Smart Chain spanning February-August 2025. Their implementation covers 11 Ethereum MEV strategies across CPMM, CLMM and Balancer V2, six strategies on BSC and two backrunning strategies on Base, with the Base scope restricted because its private mempool prevents observation of pending transactions needed by sandwich-style workflows.supported
Basis: reported finding
- supportsEVAge: Autonomous MEV Generation and Adaptation via Multi-Agent Harnessprimary disclosure
Introduction 'Prototype implementation and comprehensive evaluation'; Section 5 'Datasets & chain selection'; Section 5.3 'Cross-Chain Adaptation'
CLM-EVAGE-CROSS-CHAINFor cross-chain adaptation, the authors migrate six CPMM reference strategies from Ethereum to BSC and two backrunning strategies to Base. They report profitable historical-replay cases after migration and decreasing marginal engineering cost as shared chain-specific infrastructure is reused; the longest reported BSC migration role runtime was 2.4 hours and SBA consumed the most total input tokens at 83 million.supported
Basis: reported finding
- supportsEVAge: Autonomous MEV Generation and Adaptation via Multi-Agent Harnessprimary disclosure
Section 5.3 'Cross-Chain Adaptation'; Figure 4 and 'Economic feasibility'
CLM-EVAGE-ATTACK-SPACEThe evaluated attack/MEV space includes six reproduced reference strategies spanning sandwich, backrun and liquidity extraction plus five generated variants: SBA-HFT, JIT-HFT, BUR-HFT, LR-HFT and MLL-HFT. The generated variants preserve their parent strategy's action ordering while replacing restricted sizing with broader HFT-style search or composing HFT sizing with the parent execution structure.supported
Basis: reported finding
- supportsEVAge: Autonomous MEV Generation and Adaptation via Multi-Agent Harnessprimary disclosure
Table 1; Section 5.1 'Reference strategies' and 'Variants'; Appendix B strategy workflows
CLM-EVAGE-FAILURE-MODESThe evaluation also records clear failure modes rather than uniform success: HFT-style ternary sizing can perform poorly on non-unimodal liquidity curves, BUR-HFT simulation success on Balancer V2 falls to 1.45% with victim-slippage violations dominating failures, and MLL-HFT can underperform its baseline when the optimization converges before the profitable upper bound.supported
Basis: reported finding
- supportsEVAge: Autonomous MEV Generation and Adaptation via Multi-Agent Harnessprimary disclosure
Section 5.2 discussion of BUR-HFT, MLL-HFT and opportunity conversion
CLM-EVAGE-LIVE-BOUNDARYEVAge's published evaluation is a controlled historical-replay study, not a reported live-mainnet attack. The authors state that experiments used isolated local forks, did not broadcast transactions, did not use credentials, caused no production financial loss, and relied only on public ledger data.supported
Basis: reported finding
- supportsEVAge: Autonomous MEV Generation and Adaptation via Multi-Agent Harnessprimary disclosure
Ethical Considerations, 'Experimental Safeguards & Zero Harm'; Introduction and Section 4 validator design
CLM-EVAGE-RUNTIME-STACKThe reported implementation uses 5,367 lines of Python for six reference MEV-bot workflows and 518 lines for the agent harness. Agents run through Codex CLI v0.144.6 using GPT-5.4 with xhigh reasoning effort; deterministic execution uses Foundry/Anvil v1.5.1, Web3.py v7.14.1, SQLite v3.45.1 and Reth v1.10.2 for the local Ethereum archive node.supported
Basis: reported finding
- supportsEVAge: Autonomous MEV Generation and Adaptation via Multi-Agent Harnessprimary disclosure
Section 4.3 'Implementation'; Section 5 'Runtime environment'
CLM-EVAGE-CROSS-PROTOCOLStarting from CPMM implementations, EVAge adapts 11 strategies to Uniswap V3 CLMM and Balancer V2 on Ethereum. The authors report low exact code reuse but about 17.52% structural similarity for CPMM-to-CLMM and 18.74% for CPMM-to-Balancer adaptations, with reusable structure concentrated primarily in the planner.supported
Basis: reported finding
- supportsEVAge: Autonomous MEV Generation and Adaptation via Multi-Agent Harnessprimary disclosure
Section 5.2 'Cross-Protocol Adaptation'; Table 3 'Code Similarities Across Protocol Adaptations'
CLM-EVAGE-FORK-VALIDATIONEVAge separates agentic code generation from deterministic execution. Candidate plans are executed against isolated Anvil forks of historical EVM state, and a Proof of Concept is accepted only from the validator's concrete execution trace when the attack flow completes with positive net revenue; agents do not directly fabricate the final PoC result.supported
Basis: reported finding
- supportsEVAge: Autonomous MEV Generation and Adaptation via Multi-Agent Harnessprimary disclosure
Introduction validation design; Section 4.1 'MEV Bot Decomposition'; Section 4.2 'System Outputs'; Appendix A.3 'Validator Runtime Simulation and Profit Calculation'
CLM-EVAGE-NOVELTY-BOUNDARYThe paper labels five generated combinations as novel MEV variants, but their novelty is compositional rather than invention of the underlying HFT optimization algorithm: the authors explicitly state that the HFT search algorithm already existed in a codebase and EVAge migrated and applied it to other MEV structures.supported
Basis: reported finding
- supportsEVAge: Autonomous MEV Generation and Adaptation via Multi-Agent Harnessprimary disclosure
Section 5.1 'Variants' and 'Analysis & Insights', including the statement that the underlying HFT search algorithm pre-existed in a codebase repository
CLM-EVAGE-OFFLINE-AGENT-ROLEThe LLM agents are not placed on the latency-critical live execution path. They generate and refine bot code offline; the resulting MEV planner is deterministic and the paper reports P90 planning latencies from sub-millisecond values to 658 ms depending on strategy/protocol, while noting that end-to-end network propagation and transaction-submission latency are outside scope.supported
Basis: reported finding
- supportsEVAge: Autonomous MEV Generation and Adaptation via Multi-Agent Harnessprimary disclosure
Abstract and Section 4 system design; Section 5.4 'Planning Latency' and feasibility discussion
CLM-EVAGE-PROFIT-ASSUMPTIONSThe paper's replayed revenue measurements should be treated as experimental upper bounds rather than realizable attacker profit. The evaluation assumes unbounded attacker capital, excludes competition and private-relay submission effects, deducts gas but does not simulate asset acquisition, on-chain conversion, flash-loan repayment or builder bribes.supported
Basis: reported finding
- supportsEVAge: Autonomous MEV Generation and Adaptation via Multi-Agent Harnessprimary disclosure
Section 5 'Assumptions'; Section 5.2 discussion of JIT-HFT capital intensity
CLM-EVAGE-RELEASE-SAFEGUARDSThe authors describe a tiered release policy: high-level architecture, the core agent reasoning framework and defensive-analysis components are presented as open, while execution components directly reusable for live mempool exploitation are stated to remain restricted. The paper links an anonymized repository for code and experimental scripts.supported
Basis: reported finding
- supportsEVAge: Autonomous MEV Generation and Adaptation via Multi-Agent Harnessprimary disclosure
Ethical Considerations, 'Harm Mitigations' and 'Publication Rationale & Tiered Release'; 'Open Science'
Implications within the documented scope
Controls and mitigations
No controls or verified fix are recorded.
Unknowns and contradictions
No explicit unknowns are recorded; inspect the scope and Claim qualifications.
Sources and citation
Material revision history
- Sep 25, 2026 · Published version · first publication · revision 41
Cite this record
DiggingBeagle. “EVAge autonomously generated and adapted MEV strategies across EVM chains.” Published by DiggingBeagle Sep 25, 2026 · Public disclosure Sep 23, 2026. https://diggingbeagle.com/cases/evage-autonomously-generated-and-adapted-mev-strategies-across-evm-chains/
Citation guidance