Authors: 0xjacobzhao and ChatGPT 5

In the research report "The Intelligent Evolution of DeFi: The Evolutionary Path from Automation to AgentFi," we systematically reviewed and compared the three stages of DeFi's intelligent development: automation tools, intent-centric copilots, and AgentFi (on-chain agents). We noted that the core capabilities of a significant number of DeFi projects currently focus on swap transactions based on "intent-driven + single atomic interactions." These interactions, because they lack continuous revenue strategies, state management, or complex execution frameworks, are more aligned with the lightweight execution model of intent-centric copilots and cannot be strictly considered AgentFi.

In our high-level vision for AgentFi's future, in addition to lending and yield farming, two of the most valuable and readily implementable use cases in the near term, swap combination strategies also offer potential. When multiple swaps are combined sequentially or conditionally, a "strategy chain" is formed, such as arbitrage or yield farming. This model requires a state machine to manage positions, conditional triggering, and multi-step automated execution, embodying AgentFi's complete closed-loop system: perception → decision → execution → rebalancing.

1. DeFi Quantitative Strategy Map and Feasibility Analysis

Traditional quantitative finance (Quantitative Finance) is centered around mathematical models, statistical methods, and algorithms. It relies on data such as historical prices, trading volume, and macroeconomic indicators for data-driven decision-making. It achieves low-latency, high-frequency, automated trading through programmatic execution, supplemented by strict risk controls (stop-loss, position management, and VaR). Its main applications include high-frequency trading (HFT), trend following and mean reversion (CTA), cross-market/cross-product arbitrage, and derivatives pricing and hedging. A mature infrastructure, exchange system, and data ecosystem have been established in traditional markets.

On-Chain Quantitative Finance (OQF) continues the logic of traditional quantitative analysis, but operates within the programmable market structure of blockchain. Its data is derived from on-chain transaction records, DEX quotes, and DeFi protocol status, and is executed within smart contracts (AMMs, lending, and derivatives protocols). Transaction costs include gas, slippage, and MEV risk, and the composability of DeFi protocols allows for the construction of automated strategy chains.

Currently, on-chain quantitative finance is still in its early stages, and due to multiple factors, it is difficult to support the implementation of complex quantitative strategies: First, in terms of market structure, insufficient liquidity depth and the lack of ultra-high-speed matching mechanisms in AMMs limit the feasibility of high-frequency and large-scale transactions; second, in terms of execution and cost, on-chain block delays and high gas fees make frequent transactions unprofitable; third, in terms of data and tools, the development and backtesting environment is imperfect, the data dimension is single, and there is a lack of multi-source information such as corporate finance and macroeconomics. Among the practical DeFi quantitative strategies, the current mainstream directions are concentrated on:

Mean Reversion/Trend Following — Making buy and sell decisions based on signals from technical indicators (such as RSI, moving averages, and Bollinger Bands);

Cross-period arbitrage — represented by protocols such as Pendle, which profits from the difference between fixed and floating income;

Market making + dynamic rebalancing — actively managing AMM liquidity ranges to earn fees;

Leveraged circular income - relying on loan agreements to improve capital utilization.

Potential future growth areas include:

The maturity of the on-chain derivatives market, especially the widespread use of options and perpetual contracts;

More efficient off-chain data access, enriching model input dimensions through decentralized oracles;

Multi-agent collaboration enables automated execution and risk balancing of multi-strategy combinations.

II. Almank Positioning and Vision: AgentFi Exploration of On-Chain Quantitative Finance

In previous Crypto AI research reports, we've featured many outstanding AgentFi projects, but most have focused on fully automated operations like intent-driven DeFi execution, lending, or liquidity management. Few teams have delved deeply into quantitative trading strategies. Currently, Almanak is practically the only project on the market explicitly focused on quantitative trading. This project focuses on codeless quantitative strategy development, providing a complete toolchain covering strategy writing (Python), deployment, execution, permission management, and vaulting. This project holds a unique position within the AgentFi space and can be considered a core example of on-chain quantitative finance.

In traditional finance, inclusive finance aims to lower barriers to entry and reach a wider audience of long-tail users. Almanak extends this concept to blockchain, aiming to make quantitative trading accessible to everyone. The platform significantly reduces capital, technical, and time costs through AI-driven agent execution. Targeting active traders, financial developers, and institutional investors in the DeFi ecosystem, the platform provides comprehensive support from strategy conception to on-chain execution. This allows ordinary users, without specialized technical backgrounds, to participate in crypto asset trading and optimize returns using fully automated, on-chain, transparent, and customizable quantitative strategies.

The Almanak platform incorporates AI multi-agent collaboration (Agentic Swarm), enabling users to quickly create, test, and deploy Python-based automated financial strategies in a code-free environment for strategy development, execution, and optimization, while ensuring a non-custodial, verifiable, and scalable execution environment. Leveraging modules such as the State Machine policy framework, Safe+Zodiac permissions management, multi-chain protocol access, and Vault asset custody, Almanak maintains institutional-grade security and scalability while significantly lowering the barrier to entry for strategy development and deployment. This report systematically analyzes its product architecture, technical features, incentive mechanisms, competitive positioning, and future development path, and explores its potential value in the fields of inclusive finance and on-chain quantitative trading.

3. Almank's Product Architecture and Technical Features

Almanak's product architecture is structured along the lines of "strategy logic → execution engine → security assurance → assetization and expansion," creating a full-stack on-chain quantitative finance system for AI agent scenarios. Within this system, the Strategies module provides a strategy development and management framework from conception to implementation, currently supporting a Python SDK and with future support for natural language generation. The Deployments module, serving as the execution engine, automates the execution of strategy logic within authorized scopes and enables adaptive optimization through AI decision-making capabilities. The Wallets module utilizes a non-custodial Safe + Zodiac architecture to ensure fund and permission security, enabling institutional-grade key management and fine-grained permission control. The Vaults module transforms strategies into tokenized financial products, leveraging standardized vault contracts (ERC-7540) to facilitate fund raising, profit distribution, and strategy sharing—making strategies fully composable and seamlessly integrated into the broader DeFi ecosystem.

1. Strategies

Almanak's strategy infrastructure covers the entire process from conception to execution, including strategy conception, creation, evaluation, optimization, deployment, and monitoring. Compared to traditional quantitative trading stacks, its design differs from traditional quantitative trading stacks in three key ways: first, it focuses on AI-driven strategy development rather than relying on manual workflows; second, it incorporates a Trusted Execution Environment (TEE) to protect the privacy of strategy alphas; and finally, it utilizes a non-custodial execution model with Safe Wallet and Zodiac permission management to ensure the security and controllability of funds and execution from the ground up.

Core Features

Python-based: Written in Python, it has highly flexible and powerful programming capabilities.

State machine architecture: can implement complex decision trees and branching logic based on market conditions.

High reliability: It runs on Almanak's dedicated infrastructure and is equipped with comprehensive monitoring and failover mechanisms.

Default privacy: All strategy codes are stored encrypted to protect users' proprietary trading logic.

Transaction logic abstraction: No need to directly handle underlying blockchain interactions, wallet management, or transaction signing.

Under this architecture, the strategy framework is based on a persistent state machine design, which completely encapsulates the on-chain interaction and execution layer. Users only need to write business logic in the Strategy component. Developers can use the SDK for highly customized Python development, or in the future use the natural language strategy generator to directly describe the goals in English - the multi-agent system will then generate code for users to review. Users have complete autonomy and can approve, reject or adjust strategies before deployment, and choose to publish them as independent strategies or Vaults. Vault can also manage permissions through whitelists, providing controlled access to entities such as institutions or liquidity funds. The strategy code is encrypted and stored by default to protect user-proprietary logic; the underlying transaction construction, signing and broadcasting are all officially maintained to ensure high reliability and consistency of execution.

In the Almanak strategy library AI KITCHEN, which is currently only open to whitelisted users, we can get a glimpse of its strategy landscape: currently launched strategies include tutorial strategies (Tutorials) and technical analysis (Technical Analysis), while internally developed strategies cover liquidity mining (Liquidity Provisioning), automated looping (Automated Looping) and custom strategies (Custom Strategy). The future roadmap plans to launch arbitrage (Arbitrage), advanced yield farming (Advanced Yield Farming) and derivatives and structured products (Derivatives & Structured Product) and other advanced strategies, reflecting the complete product evolution path from basic entry to professional quantitative, from single strategy to cross-protocol complex combination.

2. Deployments

Deployments are the core execution layer connecting strategy logic with on-chain execution, responsible for automating transactions and operations within the scope of user-authorized permissions. The current primary form is StrategyDeployment, which runs on a schedule or trigger basis according to pre-defined logic. It is suitable for executing trading strategies with clear and reproducible logic, emphasizing stability and controllability. The upcoming LLMDeployment will incorporate one or more Large Language Models (LLMs) as decision engines, enabling strategies to adapt to market changes and continuously learn and optimize, exploring new trading opportunities within a strictly controlled framework.

The Deployment workflow encompasses the entire process, from authentication and authorization, policy execution, transaction construction, permission verification, to signature submission and execution monitoring. The underlying execution is handled by officially maintained core classes: TransactionManager converts policy actions into compliant on-chain transactions and simulates verification; AccountManager generates transaction signatures; and ExecutionManager broadcasts transactions, tracks status, and retries when necessary, forming a highly reliable closed loop from policy to on-chain execution. In the future, Almanak will expand to include multi-Deployment collaboration, cross-chain execution, and enhanced analytical capabilities, supporting more complex multi-agent policy operations.

3. Wallet System and Security Mechanism (Wallets)

The wallet system is central to ensuring fund security and controllable policy execution. Almanak utilizes a non-custodial solution combining Safe and Zodiac, ensuring full user ownership of funds and precisely and controllably delegating the permissions required for policy execution to the automated execution account (Deployment EOA). Users directly control their Safe Wallet through their User Wallet (EOA or ERC-4337 Smart Account). Safe Wallet incorporates the Zodiac Roles Modifier module, allowing for strict function whitelisting and parameter restrictions within the Deployment EOA, ensuring only permitted actions are executed, with permissions that can be revoked at any time.

Deployment EOAs are hosted by the platform, with their private keys stored statically using enterprise-grade encryption and managed by Google's secure infrastructure, making them inaccessible to anyone. In extreme cases, the platform will immediately notify the user to revoke permissions and generate a new EOA to replace them, ensuring uninterrupted strategy execution. To ensure continuous strategy execution, users must purchase the Autonomous Execution Fees service package, which covers on-chain operating costs (including gas). This architecture achieves institutional-grade security standards through complete isolation of funds and execution rights, refined permission management, institutional-grade key security, and rapid exception response, laying a foundation of trust for the large-scale adoption of automated DeFi strategies.

4. On-chain quantitative strategy vaults

Almanak Vaults are user-deployable, fully on-chain, permissionless vault contracts that transform trading strategies into tokenized, composable financial products. Unlike static "closed containers," these vaults are built on the ERC-7540 standard, an asynchronous extension of ERC-4626. Designed as programmable capital allocators, they integrate natively into the DeFi ecosystem.

By tokenizing AI-generated strategies, Jinku introduces a new DeFi primitive: the strategies themselves become ERC-20 assets that can be used for LPs, collateralized, traded, transferred, or combined into structured products. This composability unlocks the "DeFi Lego" at the strategy level, enabling seamless integration with protocols, funds, and structured products.

Vaults can be owned by individual curators or the community. Almanak Vaults are implemented based on Lagoon Finance's open-source contracts (MIT licensed), inheriting Lagoon's auditing and security guarantees and complying with the ERC-7540 standard. Their permission management mechanism is consistent with Almanak Wallets, relying on the Zodiac Roles Modifier to enforce function whitelisting and parameter restrictions, ensuring that all operations are strictly within authorized scope.

The operation process includes:

Strategy Binding – Bind existing Python strategies or AI-generated strategies to the vault;

Fundraising – Investors purchase Treasury tokens to gain pro rata ownership;

On-chain execution and rebalancing – the vault trades according to the strategy logic and adjusts positions dynamically;

Profit Distribution – Profits will be distributed in proportion to the token holdings, with management fees and performance fees automatically deducted.

Core advantages:

Each vault position exists in the form of an ERC-20 token, ensuring portability and interoperability;

The strategy is deterministic, auditable, and executed on-chain;

Capital is both safe and liquid – safety and composability are no longer mutually exclusive.

Developers can integrate Treasury tokens into their own protocols without permission, and fund allocators can flexibly allocate capital within the ecosystem.

In short, Almanak Vaults evolves DeFi capital management from isolated packaging containers into intelligent, composable systems. By transforming AI-generated strategies into tokenized financial primitives, it propels DeFi beyond passive yield containers toward responsive, modular capital networks, realizing the long-held vision of programmable, interoperable finance.

5. DeFi Agentic Swarm

As a one-stop platform covering the entire strategy development cycle, the Almanak AI Swarm architecture enables the independent research, testing, creation, and deployment of complex DeFi strategies, with full user control and non-custodial asset management. It aims to emulate and replace the entire process of traditional quantitative trading teams. It's worth noting that the AI Swarm "team" is composed entirely of AI agents, not real people.

Strategy Team: Converts users' natural language instructions into directly deployable on-chain strategies. This team includes strategists (designing logic), programmers (writing smart contract code), auditors (verifying correctness), debuggers (fixing errors), quality engineers (running simulation tests), permission administrators (configuring execution permissions), UI designers (building visual dashboards), and deployers (performing mainnet deployment), ensuring a complete chain from conception to implementation.

The strategy team uses LangGraph to implement deterministic process orchestration, persistent state sharing (TeamState), human-in-the-loop (HITL) verification, parallel processing, and interrupt recovery. The entire process can be automated, but manual verification is enabled by default to ensure reliability.

Alpha Seeking Team: Continuously scans the entire DeFi market, identifies market inefficiencies, explores new ideas and Alpha opportunities, and presents new logic and strategic concepts to the strategy team.

Optimization Team: We rigorously evaluate strategy performance through large-scale simulations of historical and forecasted market data. We conduct hypothetical stress testing and cyclical performance analysis before deployment to identify potential drawdowns and vulnerabilities, ensuring the stability and robustness of strategies in diverse market environments.

In addition, auxiliary AI tools include Stack Expert AI and Troubleshooting AI: the former focuses on answering various questions users may have about the Almanak technology stack and platform operations, providing real-time technical support; the latter focuses on real-time monitoring and problem location during the strategy operation process, ensuring the stability and continuity of strategy execution.

Almanak's core principles are: all AI operations will be recorded, reviewed and structured, no AI will run independently, and all strategy logic must undergo complete human-machine dual verification before going online, and users have ultimate control and custody rights.

4. Almank Product Progress and Development Roadmap Autonomous Liquidity USD Vault

Almanak has officially launched the Autonomous Liquidity USD (alUSDC Vault), a stablecoin yield-optimizing vault deployed on the Ethereum mainnet through the community. Like other lending yield-enhancing AgentFi products like Giza and Axal, its core is the Stable Rotator Agent, which continuously scans the DeFi ecosystem, identifies and captures the highest-yielding opportunities, and automatically rebalances the portfolio based on customizable risk parameters. Prior to execution, the strategy performs intelligent transaction cost analysis, adjusting positions only when the return increase is sufficient to cover all costs. Advanced routing optimization and automatic compounding maximize capital efficiency. The vault currently supports a variety of USDC derivative assets from protocols such as Aave v3, Compound v3, Fluid, Euler v2, Morpho Blue, and Yearn v3.

Almanak Liquidity Strategy and Swap Trading Strategy

Almanak's available strategies can be divided into two categories: LP series (Dynamic LP Blue Chip, Dynamic LP Degen) and indicator spot strategies (MyAmazingStrat, PENDLERSI_Momentum, VIRTUALBollingerBandsMeanReversion). Detailed strategy content is shown in the table below:

By analyzing the above strategy code with AI tools, we can draw the following conclusions:

LP dynamic market making (Blue Chip/Degen) is suitable for pools with continuous trading volume and can accept impermanent loss; Blue Chip pursues steady-state fee collection, while Degen pursues higher-frequency profit capture.

The spot indicators (EMA/RSI/BB) are lightweight and can be used for multi-parameter grid experiments with multiple indicators; however, strict cooling/slippage/minimum trading volume control is required, and attention must be paid to the pool depth and MEV.

Funding and Scale: Long-tail assets (Degen/ANIME/VIRTUAL) are more suitable for small-scale/multi-instance strategies to diversify risk control; blue-chip LPs are more suitable for medium- and long-term/higher TVL.

The minimum viable combination for real trading: Dynamic LP Blue Chip (steady-state fee collection) + RSI/BB type spot strategy (volatility capture) + small deployment of Degen LP or EMA crossover as a "high volatility test field".

Almanak Development Roadmap

Almanak's platform evolution is divided into three stages, gradually achieving expansion from a technical foundation to full-link popularization.

Phase 1 focused on infrastructure and early community building, launching a public beta version covering the complete quantitative trading stack. The Legion platform was used to establish a core user base and distribute private beta access. Simultaneously, funds were accepted into the AI-designed community treasury strategy, enabling automated on-chain management of the first batch of assets.

Phase 2 Almanak plans to fully open AI Swarm functionality to the public by the end of this year. Prior to this, access will be gradually relaxed as the system scales to large-scale use. This phase will be the official implementation of the platform's token economy and incentive system.

Phase 3 will focus on attracting global retail users, launching products tailored to savings and retirement accounts, and integrating with centralized exchanges (such as Binance and Bybit) to seamlessly integrate CeFi and DeFi. Furthermore, the platform will leverage low-risk, high-volume RWA strategies to expand asset classes and launch mobile applications, further lowering the barrier to entry for users.

In addition, Almanak will continue to expand multi-chain support (including Solana, Hyperliquid, Avalanche, Optimism, etc.), integrate more DeFi protocols, and introduce a multi-agent collaboration system (Multi-Agent System) and a trusted execution environment (TEE). Through AI-driven automatic discovery of Alpha, it will build the world's most comprehensive AI DeFi smart execution network.

5. Almank’s Token Economics and Points Incentives

Almanak's token economy aims to build an efficient and sustainable value exchange network for AI-driven financial strategies, effectively matching high-quality strategies with liquid capital on-chain. The platform creates a closed ecosystem through two core roles: Strategy & Vault Curators and Liquidity Providers. Strategy & Vault Curators utilize an agentic AI swarm to design, optimize, and manage verifiable, deterministic strategies in a codeless environment. They attract external capital by deploying permissionless vaults and collect management fees and performance sharing. Liquidity Providers deposit funds into these vaults to gain tokenized strategy exposure and participate in profit sharing.

In terms of policy privacy, policy curators can choose a private mode (closed source, not included in the policy library, and accessible only to themselves) or a public mode (open source and included in the policy library for use by the community and third-party protocols), thereby striking a balance between IP protection and knowledge sharing.

The token economic model draws on the dynamic capital allocation logic of traditional hedge funds, and integrates Bittensor's demand-driven emission allocation mechanism and Curve Finance's governance incentive model: the former allocates emissions through a weighted approach of TVL and strategy yield, encouraging capital to concentrate on high-performance strategies; the latter introduces the "veToken + Bribe" model, allowing protocol parties to increase the emission multiplier of a specific Vault by voting, thereby guiding AI agent traffic to the designated protocol.

Emissions allocation utilizes a weighted formula based on AUM and ROI, ensuring that a Vault's contribution to attracting capital and generating returns is directly converted into token rewards. A governance bonus mechanism (Almanak Wars) can grant up to 3x weight to targeted Vaults, creating an incentive market for project owners, curators, and liquidity providers to compete. To maintain long-term sustainability, a portion of protocol fees (Vault fee commissions, computing resource markups, etc.) will be returned to the emission pool, gradually offsetting the increased emission pressure as the ecosystem matures.

Token functions encompass staking and governance: Token holders can stake to obtain discounts on platform computing resources, increase voting weight, direct emissions to specific vaults, and participate in DAO governance, determining key parameters such as computing resource markups, vault fee ratios, and emission curves. The governance structure is expected to include an Ecosystem Parameters Committee and an Innovation and Development Committee, responsible for protocol parameters, revenue distribution and fund management, and ecosystem funding, respectively.

Almanak has established token distribution groups for the team (54-month linear vesting, 12-month lockup), institutional investors (48-month linear vesting, 12-month lockup), advisors (48-month linear vesting, 12-month lockup), innovation and ecosystem development (35% vesting via TGE, followed by linear vesting over the remaining 36 months), community and early adopters (45% vesting via TGE), and the Legion Community Round (two rounds: 30% TGE in the first round and 24-month linear vesting, followed by 100% TGE in the second round). The Emission Pool is used to reward network participants and early incentivize early adopters, and is distributed according to an annual halving inflation model to maintain long-term incentives and governance activity.

Almanak Points: A quantified incentive mechanism for platform participation

Almanak Points are a core mechanism for rewarding users for their engagement and contributions on the platform. This system aims to drive asset accumulation, strategic adoption, and community growth. Points are distributed in phases, each season, with adjustments to the amount of points awarded, eligible activities, and calculation method.

Users can earn Points in a variety of ways: ① Deposit funds into the community vault listed on the Almanak platform, where points are calculated based on the deposit size and holding time (currently, Vault deposits receive a 2x points multiplier); ② Hold assets in the Almanak Wallet, where points are accumulated based on the balance and holding time; ③ Enable and actively manage Deployments, where additional points are earned based on the size, duration, and complexity of assets under management; and ④ Refer new users to platform activities, where referrals receive an additional 20% of the referral's points. Points are non-transferable and non-tradable, but will be converted into tokens at a 1:1 ratio upon token issuance. Points also serve as the basis for governance rights, feature access, and ecosystem benefits.

Project financing and coin issuance strategy

Almanak has collaborated deeply with Cookie.fun and Legion.cc to introduce the Snaps/cSnaps mechanism. By analyzing the trackable contribution points on the chain through the Attention Capital Formation model, it directly links users' multi-dimensional activity in community communication, content interaction, financial support, and other dimensions with token distribution, realizing a transparent and structured distribution logic of "contribution equals ownership."

In terms of capital background, Almanak received incubation support from Delphi Digital and NEAR Foundation in its early stages, and has successively introduced investments from well-known institutions such as Hashkey Capital, Bankless Ventures, Matrix Partners, RockawayX, AppWorks, Artemis Capital, and SParkle Ventures, with a total financing amount of US$8.45 million.

January 8, 2025: Completed $1M IDO through Legion, with a valuation of $43M. 30% was unlocked through TGE, and the remaining 70% was set to a 6-month lock-up period and released linearly over 24 months.

August 21, 2025: Legion will launch its community funding round, with a valuation of $90 million USD and a target of $2 million USD, capped at $2.5 million USD. This round is open to all verified accounts, and tokens will be fully unlocked at the TGE, expected to take place in late September or early October.

Cookie DAO priority rights: The top 25 Snappers and the top 50 cSnappers will have priority to invest in $75 million in FDV, and will also be fully unlocked at the TGE.

Campaign Incentives – 0.55% of the total supply will be allocated for incentives: 0.4% to the top 500 cSnappers (80% of the reward pool), 0.1% to the top 250 Snappers, and 0.05% to $COOKIE stakers.

This issuance mechanism not only optimizes the fairness of token distribution and the participation threshold, but also deeply integrates capital fundraising, community building and long-term governance to form a sustainable community of interests, laying the foundation for Almanak's long-term expansion in the AI × DeFi track.

VI. Investment Logic and Potential Risk Analysis Investment Logic

Almanak currently aligns more closely with its positioning as the "easiest-to-use retail DeFi strategy sandbox." Its significant advantages in user experience, security architecture, and low barrier to entry make it particularly suitable for retail investors with no coding experience or limited on-chain strategy experience. Almanak's core competitiveness lies in its deep integration of an AI multi-agent architecture with a non-custodial execution system, providing institutional-grade security and policy privacy without sacrificing performance. Its technology stack, comprised of a TEE (Trusted Execution Environment) + Safe Wallet + Zodiac Roles Modifier, enables fully automated on-chain execution with permission management down to the granularity of contract function parameters, significantly outperforming most AgentFi models that rely solely on EOA signatures.

The technical architecture forms a complete closed loop: from data acquisition (Sensors), strategy logic execution (persistent state machine architecture, Prepare/Validate/Sadflow), transaction execution (TransactionManager/AccountManager/ExecutionManager), monitoring and indicator systems, to productization (ERC-7540 Vault) and external fundraising and fee collection, the chain is smooth and commercially scalable. In particular, the Vault's productization capabilities enable strategies to be directly upgraded from self-use tools to publicly issueable financial products, generating scalable management fee and performance-sharing revenue for the platform.

On the operational side, the Points incentive system has been launched. The rules are transparent and guided by AUAM (Assets under Agentic Management), which can effectively drive the locked-in volume and activity. By analyzing the traceable contribution points on the chain through the Attention Capital Formation model, capital raising, community building and long-term governance are deeply integrated to form a sustainable community of interests, laying the foundation for Almanak's long-term expansion in the AgentFi track.

Potential risks

Although Almanak has a relatively high degree of technical and functional system completion, it still faces several key challenges:

First, protocol testing is not yet fully open. Currently, it only supports lending protocols on the Ethereum mainnet, as well as LP and swap strategies based on Uniswap V3. The underlying Kitchen framework already supports Ethereum, Base, and Arbitrum, and has the capability to expand to 8+ EVM chains and 200+ protocols. The pace of wider openness, including multi-chain expansion (such as Base, Solana, and Hyperliquid), multi-protocol integration, and CEX integration, will directly impact strategy diversity, profit opportunities, and market competitiveness.

Secondly, the strategy level is still relatively basic. Existing strategies primarily rely on entry-level technical analysis (TA), which still lags behind institutional or professional quantitative standards. Future developments require the introduction of a richer library of advanced strategies (covering on-chain liquidity management, funding rate arbitrage, cross-pool spreads, and multi-signal integration), as well as improved backtesting and simulated trading tools, a built-in cost optimization engine, and multi-strategy portfolio management capabilities. Only then can it become the preferred entry point for retail investors into DeFi quantitative trading.

Furthermore, the ecosystem and user base are still in their early stages. Although the points program and Vault mechanism have been launched and have entered the first tier of AgentFi, the TVL growth of AI-driven Vault management still needs time to be verified, and user activity and retention will be key indicators in the medium and long term.

Overall, Almanak boasts a mature technical architecture, a clear commercialization path, and a strong incentive mechanism, making it a rare asset in the AI-driven on-chain quantitative and asset management space. However, the speed of its ecosystem expansion, the evolution of the competitive landscape, and the stability of its technological implementation will be key factors in determining its long-term leadership.