Recent advances in distributed system design have shifted from centralized client–server models to decentralized frameworks that enable broad inter-organizational collaboration. While Decentralized Autonomous Organizations (DAOs) leverage blockchain for governance and token-based transactions, existing platforms face two major limitations: the absence of decentralized authorization mechanisms and inadequate support for large-scale data management. This paper introduces Agentic DAO, a next-generation DAO architecture that integrates multi-agent authorization and scalable data governance into a blockchain-centric framework. The platform employs a Trust Scoring Agent using Graph Neural Networks (GNNs) to predict transactional reliability and a Data Management Agent to ensure consistency between blockchain and IPFS-based storage. Governance is reinforced through SoulBound Token (SBT)-based credentials to prevent Sybil attacks and enforce fair participation. A prototype implementation demonstrates the feasibility of trust-based access control and robust data synchronization. Experimental results confirm that Agentic DAO enhances transparency, scalability, and fairness compared to conventional DAO models, enabling secure and collaborative engagement in decentralized ecosystems.
The evolution of the World Wide Web is conventionally described in three eras-Web1 (read), Web2 (read-write), and Web3 (read-write-own)-each of which relocated a distinct form of power along the value chain: from publishers, to platforms, to ledger-anchored token holders. This paper introduces and defends a fourth: Web4 (read-write-own-act), in which sovereign ownership is not confined to tokens or ledger records but is co-specified across every operational plane of the network-identity, physical connectivity, compute, naming, developer tooling, application delivery, and consumer hardware-and in which autonomous agents act on behalf of that ownership under cryptographic, time-boxed capability grants. The central original contribution is the Ownership Thesis: the axiom that every layer of the stack must be reducible to a claim controlled by a decentralized identifier the subject holds directly, not one that is issued or held by an operator on the subject's behalf. From this axiom we derive a seven-layer reference architecture and formalize the Synthetic Web as the class of network activity structurally defined by agent-executed action under such capability grants. We position each layer against the relevant standards (W3C DID Core v1.1, W3C Verifiable Credentials Data Model v2.0, Model Context Protocol, Agent-to-Agent v1.0.0, NIST CAISI AI Agent Standards Initiative) and provide a security analysis covering key management, alias unlinkability via BBS+ selective disclosure, Sybil resistance, delegated-authority scope creep, and data-sovereignty enforcement, with explicit identification of open problems that require further verification.
Memory-enabled large language model (LLM) agents, particularly those deployed in long-horizon, tool-using settings such as Web3-style autonomous workflows, introduce security risks that extend beyond single-prompt injection. By persisting and reusing information across interaction steps and sessions, these agents enable memory poisoning attacks in which adversarial inputs modify persistent agent state and influence future decisions after benign intermediate interactions. Recent work on context manipulation and “fake memories” demonstrates that adversarial content can be injected into an agent’s prompt-visible inputs or persistent memory; however, existing evaluations largely analyze such attacks at isolated interaction steps or static context snapshots, obscuring their temporal dynamics. In this paper, we present the first large-scale, trajectory-level measurement framework for analyzing temporal memory poisoning in memory-enabled LLM agents. We construct a schema-constrained dataset of 2,614 multi-step attack trajectories spanning four attack families,chain poisoning, policy rewriting, backdoor triggering, andslow drift, executed over shared persistent memory. We define temporal risk metrics over multi-step interaction trajectories that capture delayed activation, non-monotonic escalation, and the earliest point at which attacks become distinguishable from benign behavior. Our empirical results show that a substantial fraction of attacks remain indistinguishable from benign behavior until late-stage activation, despite exhibiting low or medium risk at all earlier steps. Slow-drift and backdoor-trigger attacks, in particular, systematically evade step-local evaluation until terminal interactions, while chain poisoning and policy rewriting exhibit non-monotonic risk trajectories. These findings demonstrate that memory poisoning risk is inherently temporal and cannot be reliably assessed using prompt-level or step-isolated evaluation, motivating trajectory-aware benchmarks for agent security.
Dockerised blockchain solution can mitigate the low levels of distributed technology adoption in small and medium enterprises. It can be done via designing and implementing an environment which inherits ease of deployment and scalability of containerized systems with safety and transparency of distributed applications. Practical implementation of a dockerized blockchain solution designed as a demonstrative implementation for existing client–server architecture is described in this paper. This solution uses Docker containers to simplify the setup and deployment of a private blockchain network, a mediator server and a reverse proxy. Implementation of this system on a low scale demonstrates feasibility of integrating blockchain technology into existing business processes without fundamental architectural changes and acknowledges deployment and maintaining challenges that usually accompany distributed systems using private blockchain. Discussed implementation is a demonstration of designed architecture being potentially a reproducible and easily maintainable environment for logging and validating data through an immutable ledger on a smaller scale. Proof of concept successfully validates the core idea. The implementation shows a mediator server intercepting client request, recording them on a private Ethereum blockchain via a JSON-RPC interface, and then forwarding them to the original server. This confirms the solution’s ability to introduce a trusted, intermediate layer for data immutability. The project demonstrates a working framework for embedding distributed ledger technologies into client–server ecosystems. While the current Proof of Work consensus mechanism presents scalability limitations, the architecture provides a strong foundation for future research, including migrating to more efficient consensus mechanisms and integrating smart contracts.
Existing agent communication frameworks face critical limitations in providing verifiable audit trails without compromising the privacy and confidentiality of agent interactions. The protection of agent communication privacy while ensuring auditability emerges as a fundamental challenge for applications requiring accurate billing, compliance verification, and accountability in regulated environments. We introduce a framework for auditing agent communications that keeps messages private while still checking they follow expected rules. It pairs zero-knowledge proofs with the existing Model Context Protocol (MCP) so messages can be verified without revealing their contents. The approach runs in lightweight networks, stays compatible with standard MCP exchanges, and adds asynchronous audit verification to confirm format and general message types without exposing specifics. The framework enables mutual audits between agents: one side can check communication content and quality while the other verifies usage metrics, all without revealing sensitive information. We formalize security goals and show that zk-MCP provides data authenticity and communication privacy, achieving efficient verification with negligible latency overhead. We fully implement the framework, including Circom-based zero-knowledge proof generation and an audit protocol integrated with MCP's bidirectional channel, and, to our knowledge, this is the first privacy-preserving audit system for agent communications that offers verifiable mutual auditing without exposing message content or compromising agent privacy.
Access control in P2P network is extremely challenging due to several reasons such as decentrailized network, anonymous connectivity, high churn, resource constraints and large attack surface. The number of attributes to be tracked grows exponentially as P2P networks scale, exacerbating issues in traditional approaches such as RBAC, ABAC, RuBAC. This paper presents a novel blockchain-based access control framework for P2P networks using Ethereum smart contracts to offer an adaptable, decentralized, and scalable solution. Our token economy framework incentivizes nodes that demonstrate consistent compliance and penalize malicious actions using fungible ERC-20 tokens. Hierarchical static role-based structure consisting of regular members, primary and secondary group is the foundational layer of role management. Dynamic state management through a finite state machine classifies peers as benign, suspicious, or malicious mapping their behavior to the STRIDE threat model. We have used standard blockchain tool stack consisting of OpenZeppelin (contract library), Mocha (testing), and Web3.js (deployment). Our framework is tested with eight test cases, such as privilege escalation and denial-of-service attacks, and built using three types of Ethereum smart contracts. Our focus is on collecting four metrics gas cost, latency, execution time, simultaneous requests that provides a comprehensive view of realistic network activity. This study goes beyond conceptual nature of state-of-art architecture to showcase the realistic mapping between resources, peer roles and standard threats in a P2P network. The experimental results shows low gas costs (upto 1.4 million gas) and fast execution times (80–550 ms), confirming the design efficiency. The framework handles up to five concurrent requests with minimal performance impact, demonstrating strong scalability across roles and threat conditions. This framework enables scalable, behavior-driven governance in P2P networks, enhancing IoT and Web3 security.
This paper introduces a novel gradient-based framework for maintaining protocol invariants in decentralized finance (DeFi) systems, representing a fundamental departure from traditional reactive monitoring approaches. Rather than detecting violations after they occur and adjusting system perceptions, this framework implements a proactive closed-loop control system that influences external market dynamics through calibrated protocol actions. The core innovation lies in tracking confidence gradients—including velocity, acceleration, and momentum—to predict potential invariant violations before they materialize. The system employs a nine-parameter calibration vector that dynamically adjusts based on confidence trends, enabling the protocol to respond appropriately to increasing (healthy), decreasing (crisis), or stable market conditions. During crisis scenarios, aggressive parameter adjustments trigger protocol actions such as enhanced lender incentives and borrower penalties, directly influencing market participant behavior to maintain critical constraints like maximum lending rates. The framework establishes mathematical foundations for confidence gradient metrics, parameter constraint spaces, and crisis level assessment. It demonstrates how calibrated algorithmic behavior generates measurable influence on external actors, effectively changing market reality rather than merely observing it. Theoretical analysis proves bounded confidence oscillations and probabilistic invariant maintenance guarantees under the proposed calibration schemes. Practical implementation considerations include smart contract architecture, gas optimization strategies through batched updates and fixed-point arithmetic, and parameter discretization for on-chain deployment. A numerical crisis response example illustrates the system's ability to prevent rate violations through coordinated supply increases and demand reductions. The paper positions this approach as foundational for the Kera Protocol ecosystem, with applications extending beyond DeFi lending to automated market makers, stablecoin protocols, governance systems, and cross-chain bridges. This paradigm shift from observation to calibrated influence represents a significant advancement in blockchain protocol stability mechanisms.
The absence of a fully decentralized, verifiable, and privacy-preserving communication protocol for autonomous agents remains a core challenge in decentralized computing. Existing systems often rely on centralized intermediaries, which reintroduce trust bottlenecks, or lack decentralized identity-resolution mechanisms, limiting persistence and cross-network interoperability. We propose the Decentralized Interstellar Agent Protocol (DIAP), a novel framework for agent identity and communication that enables persistent, verifiable, and trustless interoperability in fully decentralized environments. DIAP binds an agent's identity to an immutable IPFS or IPNS content identifier and uses zero-knowledge proofs (ZKP) to dynamically and statelessly prove ownership, removing the need for record updates. We present a Rust SDK that integrates Noir (for zero-knowledge proofs), DID-Key, IPFS, and a hybrid peer-to-peer stack combining Libp2p GossipSub for discovery and Iroh for high-performance, QUIC based data exchange. DIAP introduces a zero-dependency ZKP deployment model through a universal proof manager and compile-time build script that embeds a precompiled Noir circuit, eliminating the need for external ZKP toolchains. This enables instant, verifiable, and privacy-preserving identity proofs. This work establishes a practical, high-performance foundation for next-generation autonomous agent ecosystems and agent-to-agent (A to A) economies.
As the "agentic web" takes shape-billions of AI agents (often LLM-powered) autonomously transacting and collaborating-trust shifts from human oversight to protocol design. In 2025, several inter-agent protocols crystallized this shift, including Google's Agent-to-Agent (A2A), Agent Payments Protocol (AP2), and Ethereum's ERC-8004 "Trustless Agents," yet their underlying trust assumptions remain under-examined. This paper presents a comparative study of trust models in inter-agent protocol design: Brief (self- or third-party verifiable claims), Claim (self-proclaimed capabilities and identity, e.g. AgentCard), Proof (cryptographic verification, including zero-knowledge proofs and trusted execution environment attestations), Stake (bonded collateral with slashing and insurance), Reputation (crowd feedback and graph-based trust signals), and Constraint (sandboxing and capability bounding). For each, we analyze assumptions, attack surfaces, and design trade-offs, with particular emphasis on LLM-specific fragilities-prompt injection, sycophancy/nudge-susceptibility, hallucination, deception, and misalignment-that render purely reputational or claim-only approaches brittle. Our findings indicate no single mechanism suffices. We argue for trustless-by-default architectures anchored in Proof and Stake to gate high-impact actions, augmented by Brief for identity and discovery and Reputation overlays for flexibility and social signals. We comparatively evaluate A2A, AP2, ERC-8004 and related historical variations in academic research under metrics spanning security, privacy, latency/cost, and social robustness (Sybil/collusion/whitewashing resistance). We conclude with hybrid trust model recommendations that mitigate reputation gaming and misinformed LLM behavior, and we distill actionable design guidelines for safer, interoperable, and scalable agent economies.
Aniket P. Kakde, Karan M. Bhoyar, Muhammad Aiman Shad, Prof. Sudesh A. Bachwani
Autonomous agents powered by Large Language Models (LLMs) require reliable and standardized frameworks to connect tools, exchange contextual information, and synchronize tasks across diverse systems. Despite growing interest in such agents, current integration with external tools remains disjointed. Developers often have to manually create interfaces, handle authentication protocols, and navigate incompatible function-calling standards across platforms. To overcome these limitations and promote the evolution of agentic AI, it is critical to establish standardized communication protocols that ensure interoperability—enabling agents and systems to seamlessly discover each other’s capabilities, share data, and coordinate operations. This paper explores a structured overview of emerging communication standards for agents, focusing on the Model Context Protocol (MCP), Agent Communication Protocol (ACP), Agent-to-Agent Protocol (A2A), and Agent Network Protocol (ANP). MCP utilizes a JSON-RPC based client-server architecture to enable secure execution of tools and well-typed data transfer. ACP introduces a REST-compliant message structure with support for asynchronous streaming and multipart formats, facilitating rich, multimodal agent outputs.A2A enables agents to delegate tasks peer-to-peer using capability-rich Agent Cards, enabling scalable and distributed workflows across organizations. ANP facilitates agent discovery and secure collaboration in open networks, leveraging decentralized identifiers (DIDs) and semantic graphs based on JSON-LD.
The growth of freelancing exposed several drawbacks in the platforms that are now in use, including high service costs, fraud risks, and late payments [1]. A well-designed system architecture is necessary to develop a freelance platform to guarantee scalability, security, and effectiveness. A modular architecture was chosen to give the required flexibility, scalability, and ease of maintenance to overcome these issues. This paper presents the design of a microservices-based architecture for a crypto freelance exchange platform, which uses Domain-Driven Design principles [2]. The architecture is built to support decentralized transactions, smart contract integration, and secure user authentication. It also ensures high availability and fault tolerance. The system uses a multi-layered architecture incorporating PostgreSQL, MongoDB, Redis, and Web3.js. The main components are a web application, KrakenD API Gateway, auth microservice, files microservice, and main microservice for managing transactions, orders, and payments. They are designed to meet critical non-functional requirements such as scalability, security, and maintainability. Each service can be independently deployed, updated, and scaled according to transaction volumes. With an emphasis on security, the platform uses JWT token techniques and multi-factor authentication to authenticate users. Also, the integration of blockchain technology enhances transparency, enabling freelancers and clients to have a trusted record of all transactions and reducing the risk of fraud. The architecture is visualized through UML and C4 model diagrams showing component interactions and service orchestration. This paper also discusses the rationale behind the chosen technologies, security mechanisms, and the benefits of using a modular microservices approach for building a crypto freelance platform.
This dissertation investigates how tokenised claims and algorithmic governance reshape interactions in Web3, with a particular focus on business-to-business (B2B) settings. Building on the insight that digital platforms and infrastructures are mutually entangled—platforms acquiring infrastructural roles and infrastructures accumulating platform logics—the study examines how this entanglement reappears in blockchain-based systems and what it means for value creation, value distribution, and institutional control. Rather than assuming decentralization as an outcome, the dissertation asks how governance is actually assembled across code, organizations, and markets, and how these assemblies channel rights, risks, and rents over time. In this sense, the thesis extends platform/infrastructure scholarship into the Web3 domain, showing how infrastructuring and platformization remain co-constitutive under new technical conditions (e.g., programmable settlement, public ledgers, composability). The research is guided by the following question: How does algorithmic governance of tokenised claims affect dynamics of value creation and distribution in Web3? The thesis addresses a gap in extant work by analysing the combined economic and governance consequences of tokenisation in commercial contexts, rather than treating governance as either purely technical (smart contracts) or purely institutional (foundations, standards, regulators). Methodologically, this research adopts a qualitative, interpretive design centred on semi-structured interviews with founders and leads of Web3 projects oriented toward commercialization and enterprise use. Interview evidence is triangulated with document analysis (white papers, governance docs, upgrade logs) to trace how decision rights are allocated, which boundary resources act as chokepoints, and how incentives and accountability are engineered. The sample focuses on projects that tokenise rights and obligations to orchestrate inter-firm exchanges (e.g., guarantees, attribution, royalties), enabling a consistent comparison of governance choices and their distributional signatures. Theoretically, the thesis contributes a layered view of Web3 governance that differentiates transaction governance (smart-contract rules that execute exchanges) from platform governance (meta-rules that structure participation, evolution, and control)—layers that are interdependent yet analytically distinct. Across cases, transaction governance supplies deterministic settlement (escrows, splits, auctions), while platform governance defines constitutional levers (eligibility schemas, listings, parameter updates, treasury policy, emergency powers). This distinction clarifies why “more on-chain” does not automatically imply “more decentralised”: instruments can be automated while decision rights remain concentrated. The framing resonates with and extends platform governance scholarship that locates governance in the ongoing division of decision rights, control mechanisms, and incentives among interdependent actors. Empirically, the thesis identifies three governance models—monocentric, moderately polycentric (P2), and highly polycentric (P1)—and analyses how each allocates rights and rents. Monocentric configurations recentre constitutional authority in a focal hub (firm, foundation, tightly bonded coalition), delivering speed, legal legibility, and coherent risk management, while concentrating surplus upstream via control of boundary resources (standards, registries, upgrade cadence, listings). Moderately polycentric arrangements disperse constitutional authority across overlapping venues (token voters, stewards, committees, standards groups), pairing automated execution at the edge with contestable meta-rules and auditable, replaceable discretion. Highly polycentric designs thin the platform layer and push coordination into markets and minimal, auditable rules (fee markets, open listings, plural oracles), improving neutrality and exit but requiring continuous work to diffuse emergent chokepoints (indices, bridges, relays). The patterns observed align with infrastructure/platform research on how control points shape innovation and value capture and with blockchain governance work emphasizing the allocation of decision and control rights. For B2B contexts, the analysis suggests a pragmatic equilibrium. Applications that demand auditability, finality, and accountable remediation (e.g., elections, trade guarantees) gravitate toward monocentric settlements; applications with heterogeneous actors and rapid iteration (e.g., creator and talent markets) benefit from moderately polycentric designs that preserve micro-level determinism with macro-level contestability. Across models, tokenisation expands what can be coordinated, but distributional outcomes hinge on who controls admission, measurement, and upgrade pathways. Accordingly, the thesis proposes design heuristics: separate transaction and platform governance, publish change logs and revocation paths, pluralise attestors at measurement junctions, time-box mandates, and keep credible exit technically and institutionally real. In sum, the dissertation advances an integrated account of Web3 as a political economy of programmable claims and layered governance. It shows how infrastructuring and platformization fold into one another under blockchain conditions, how distinct governance models redistribute rights and rents, and how B2B value propositions depend as much on constitutional design as on code. The framework equips scholars and practitioners to evaluate Web3 systems not by decentralisation rhetoric, but by the concrete allocation of decision rights, boundary resources, and incentives across layers and venues.
Ovaj rad predstavlja razvoj prototipa decentralizirane društvene mreže temeljene na tehnologijama Web3. S obzirom na sve veće nepovjerenje korisnika prema tradicionalnim, centraliziranim društvenim mrežama, cilj je izraditi rješenje koje korisnicima omogućuje privatnost, sigurnost i vlasništvo nad vlastitim podacima. Aplikacija koristi blockchain Ethereum za upravljanje identitetom i interakcijama korisnika i Metamask za autentikaciju. Sadržaj se pohranjuje distribuirano putem sustava IPFS i mreže istorazinskih dionika Gun. Implementirane su funkcionalnosti poput objavljivanja sadržaja, spremanje objava, označavanje objava sa "sviđa mi se", komentiranja, slanja zahtjeva za prijateljstvo i drugih interakcija. Uz sve navedeno, aplikacija uključuje izravni (peer-to-peer) chat, te sustav za nagrađivanje korisnika putem ERC-20 tokena za aktivnosti, interakciju i kvalitetan sadržaj. Sustav je implementiran lokalno pomoću IPFS i GUN čvorova na dva uređaja. Postavljen je temelj za daljnju optimizaciju i širenje funkcionalnosti, kako bi se ostvarila stvarna, sigurna i decentralizirana mreža.
Ovaj rad prikazuje dizajn i implementaciju decentralizirane aplikacije za kvizove koja spaja prednosti blockchain tehnologije s praktičnošću Web2 tehnologija. Koristeći hibridnu arhitekturu on-chain i off-chain obrade podataka. Korišteni su Angular, Ethers.js, MetaMask na frontendu, Node.js/Express i MongoDB na backendu te Solidity pametni ugovori na ethereum virtual machine (EVM)-kompatibilnoj mreži (Polygon Amoy za testiranje). Arhitektura uključuje QuizFactory za kreiranje ins- tanci, osnovni Quiz za besplatne sesije i proširivi QuizWithFee koncept za nagradni fond. Testiranjem je utvrđena funkcionalnost sustava, ali i ukazan je linearni porast troškova s porastom broja igrača i pitanja. Rezultat je na kraju funkcionalna apli- kacija za live kvizeve u kojem sudionici nemaju on-chain troškove, ključni ishodi su javno provjerljivi na blockchain-u, a visoko-frekventne interakcije ostaju izvan lanca radi boljeg korisničkog iskustva i nižih troškova; sustav je modula
This research article presents a novel architecture to empower multi-agent economies by addressing two critical limitations of the emerging Agent2Agent (A2A) communication protocol: decentralized agent discoverability and agent-to-agent micropayments. By integrating distributed ledger technology (DLT), this architecture enables tamper-proof, on-chain publishing of AgentCards as smart contracts, providing secure and verifiable agent identities. The architecture further extends A2A with the x402 open standard, facilitating blockchain-agnostic, HTTP-based micropayments via the HTTP 402 status code. This enables autonomous agents to seamlessly discover, authenticate, and compensate each other across organizational boundaries. This work further presents a comprehensive technical implementation and evaluation, demonstrating the feasibility of DLT-based agent discovery and micropayments. The proposed approach lays the groundwork for secure, scalable, and economically viable multi-agent ecosystems, advancing the field of agentic AI toward trusted, autonomous economic interactions.
The fragmentation of AI agent ecosystems has created urgent demands for interoperability, trust, and economic coordination that current protocols -- including MCP (Hou et al., 2025), A2A (Habler et al., 2025), ACP (Liu et al., 2025), and Cisco's AGP (Edwards, 2025) -- cannot address at scale. We present the Nanda Unified Architecture, a decentralized framework built around three core innovations: fast DID-based agent discovery through distributed registries, semantic agent cards with verifiable credentials and composability profiles, and a dynamic trust layer that integrates behavioral attestations with policy compliance. The system introduces X42/H42 micropayments for economic coordination and MAESTRO, a security framework incorporating Synergetics' patented AgentTalk protocol (US Patent 12,244,584 B1) and secure containerization. Real-world deployments demonstrate 99.9 percent compliance in healthcare applications and substantial monthly transaction volumes with strong privacy guarantees. By unifying MIT's trust research with production deployments from Cisco and Synergetics, we show how cryptographic proofs and policy-as-code transform agents into trust-anchored participants in a decentralized economy (Lakshmanan, 2025; Sha, 2025). The result enables a globally interoperable Internet of Agents where trust becomes the native currency of collaboration across both enterprise and Web3 ecosystems.
Abstract Pixel-Web3 Wallet is a hierarchical deterministic (HD) wallet designed for secure and decentralized asset management across multiple blockchain networks, including Ethereum and Solana. Unlike traditional wallets that depend on browser extensions or centralized servers, Pixel offers a web-based solution with user-controlled security through locally stored seed phrases. This paper explores the wallet’s architecture, security framework, and innovative features, such as real-time balance updates and flexible recovery options. Additionally, the research evaluates the scalability of Pixel and its potential expansion to support more blockchain networks. By eliminating reliance on third-party services, Pixel enhances accessibility while maintaining strong security, making it a promising solution for blockchain enthusiasts, traders, and developers. Keywords: Blockchain, HD Wallet, Cryptocurrency,Web3,Ethereum,Solana, Security
Smart contracts power decentralized financial (DeFi) services but are vulnerable to security exploits that can lead to significant financial losses. Existing security measures often fail to adequately protect these contracts due to the composability of DeFi protocols and the increasing sophistication of attacks. Through a large-scale empirical study of historical transactions from the 37 hacked DeFi protocols, we discovered that while benign transactions typically exhibit a limited number of unique control flows, in stark contrast, attack transactions consistently introduce novel, previously unobserved control flows. Building on these insights, we developed CrossGuard, a novel framework that enforces control flow integrity onchain to secure smart contracts. Crucially, CrossGuard does not require prior knowledge of specific hacks. Instead, configured only once at deployment, it enforces control flow whitelisting policies and applies simplification heuristics at runtime. This approach monitors and prevents potential attacks by reverting all transactions that do not adhere to the established control flow whitelisting rules. Our evaluation demonstrates that CrossGuard effectively blocks 35 of the 37 analyzed attacks when configured only once at contract deployment, maintaining a low false positive rate of 0.26% and minimal additional gas costs. These results underscore the efficacy of applying control flow integrity to smart contracts, significantly enhancing security beyond traditional methods and addressing the evolving threat landscape in the DeFi ecosystem.
This paper introduces Bank-Guard Fusion System, a decentralized blockchain wallet system designed for secure cryptocurrency management. Utilizing blockchain's decentralized, cryptographic strengths, the platform enables wallet creation, balance checks, and transactions in a user- friendly, secure environment. Built with React.js, Flask, and MongoDB, it provides a seamless interface and decentralized data handling. Integrating Proof of Work (PoW) and Proof of Stake (PoS) for transaction validation, this solution enhances trust, transparency, and security, bridging users and blockchain technology to facilitate sustainable, scalable, and robust digital transactions. Keywords – Blockchain, Cryptocurrency, Flask, MongoDB, React.js, Decentralized Application, Wallet, Secure Transactions, PoW, PoS, XMSS, Hybrid Security
This thesis investigates the security challenges of blockchain consensus mechanisms, arguing that sub-stratal to these challenges is the widely adopted mindset of probabilistic finality, which in itself is an inherent and significant vulnerability. Core to this problem is the inability or probabilistic protocols such a Proof of Work and Proof of Stake to provide the necessary deterministic and irreversible transaction settlement that high-integrity systems require. This is primarily due to their use of a “longest chain” rule. This deficiency is immensely amplified on new blockchains, as they do not possess the accumulated economic and computational security of mature networks. New blockchains face a serious “bootstrap security problem” which makes them highly prone to catastrophic 51% attacks. The research methodology follows a multi-stage, systematic approach, beginning with a comprehensive literature review to classify and analyse the existing threat landscape across major consensus protocols. Then a targeted theoretical critique of probabilistic finality is presented next, as is a systematic evidence-based assessment of some of the economic and architectural vulnerabilities of nascent networks. The findings of this analytical work confirm that the security of a blockchain is not static but evolves, and that the initial phase of a network's lifecycle is its most insecure. The main contribution of this thesis is the design and proposal of a novel consensus algorithm called “Erdos” that aims to address these issues directly. Erdos offers a deterministic process for block finalization that involves multiple rounds of vetting and a fair, resource-independent node selection. The design provides a strong fork resistance that stops the risk of 51% attacks and double spending right from the inception and launch of a network. Moreover, it reduces the centralisation pressure that proves to be a problem for nascent protocols. The primary implication of this research is that the future of secure, decentralised systems, particularly for new and emerging applications, necessitates an architectural shift away from probabilistic models towards deterministic, fair, and resilient consensus protocols as exemplified by the Erdos algorithm.
Consensus serves as a foundational mechanism in both social coordination and distributed technical systems. While machine consensus research in engineering focuses on fault tolerance and synchronization, social science emphasizes human deliberation, participation, and governance. However, the increasing convergence of human and machine decision making, exemplified by decentralized autonomous organizations (DAOs), intelligent agents, and cyber-physical social systems, demands a more integrated and theoretically robust understanding of consensus. This thesis addresses this interdisciplinary gap by investigating consensus across three interconnected dimensions: probabilistic fault-tolerant consensus systems, human-driven voting mechanisms in DAOs, and a unified conceptual framework bridging human and machine consensus. The first part of the thesis focuses on distributed fault-tolerant consensus in uncertain environments. Traditional approaches often rely on deterministic assumptions about node failures and fixed quorum rules. These assumptions may fail to reflect real-world systems where node behaviour is influenced by heterogeneous reliability and probabilistic failures. To address this limitation, a probabilistic modelling framework is proposed, treating node reliability as a stochastic variable. Within this framework, consensus outcomes are classified into three categories: safe, risky, and compromised. A new concept, referred to as the reliability quorum, is introduced to provide a more flexible threshold for achieving consensus based on targeted reliability levels. This model enables system designers to tailor fault tolerance according to specific reliability requirements, providing both analytical clarity and practical adaptability. The second part investigates consensus in decentralized systems primarily driven by human-oriented agents, using DAO voting as a representative case. In contrast to deterministic coordination among machines, DAO consensus arises from voluntary participation, heterogeneous voting power, and non-uniform approval conditions. To guide the analysis, the thesis introduces the DAO governance triangle alongside the SEED framework, which qualitatively evaluates voting mechanisms across four dimensions: Security, Efficiency, Effectiveness, and Decentralization. Building on this conceptual foundation, the study proceeds to a quantitative investigation of two key SEED dimensions. For decentralization, a stochastic process model is proposed to capture probabilistic participation and power distribution, leading to the formulation of the Consistency Rate and the Decentralization Coefficient as quantitative indicators. For efficiency, the model is further extended to characterize the interactions among participation probability, voting duration, and approval rate, enabling a formal evaluation of voting responsiveness and resource usage. Simulation results support both aspects of the analysis, revealing how power concentration, turnout behaviour, and mechanism design jointly influence decentralization and efficiency in DAO voting. The third part presents a unifying conceptual framework to analyse consensus across human, machine, and human-machine hybrid systems. Despite disciplinary differences, the thesis identifies three core components of any consensus process: participants (the actors of agreement), communication (the medium of exchange), and state (the evolving representation of agreement). Framing consensus as an entropy-reduction process that resolves cognitive or informational divergence, this abstraction enables comparative analysis across diverse systems. The framework also distinguishes among human consensus, machine consensus, and human-machine hybrid consensus, and offers design guidelines aligned with the characteristics and limitations of each. Together, these three threads construct a comprehensive theory of consensus that connects distributed computation, social governance, and emerging hybrid collectives. By integrating modelling, evaluation, and abstraction, this thesis contributes a multi-layered foundation for understanding and designing consensus mechanisms that are robust, scalable, and trustworthy in increasingly decentralized and intelligent environments.