Bridge SoK: ToDos

Target: COMST, 160+ citation 1. Introduction 4. Preliminaries 6. Formalization 7. **Bridge solution** - cross-reference formalization and preliminaries 8. **Bridge attack** - internal logic of listing the attacks (root causes) - cross-reference formalization! - add empirical attacks 9. **Defense solutions** - cross-reference formalization! - empirical defense - insights / judements / assesments 1. Discussion - limitation of this particular paper - disclaimers / caveats of the entire paper (limitation in data collection) - future work 2. literature review / related works - highlight inconsistencies / drawbacks with existing literature - insert literature comparison table here - "bridging the gap": gap between industry (real life / emprical incidents) and literature (subjects studied) 11. Conclusion ## Next Steps For Paper - Updated 1. - [ ] Introduction - [ ] a. Changes to structure --- 2. - [ ] Bridge Solutions - [x] a. Introducing Popular CCB Protocols - [ ] b. Connectivity in the Global Blockchain Space - [ ] c. Add in a new table which shows all the blockchains and weather they’re connected or not - [x] d. cross-reference formalization and preliminaries --- 3. - [ ] Bridge attack - [ ] a. Add empirical attacks - [ ] b. Internal logic of listing the attacks (root causes) - [ ] c. cross-reference formalization --- 4. - [ ] Defence solutions - [ ] a. Add empirical defence - [ ] b. cross-reference formalization --- 5. - [ ] Discussion - [ ] a. Limitations of the paper - limitation in data collection - [ ] b. Future work --- 6. - [ ] Literature review / related works - [ ] a. Inconsistencies / drawbacks with existing literature - [ ] b. literature comparison table - [ ] c. Gap between industry and literature --- 6. - [ ] Update Conclusion --- # Dissecting Attacks and Vulnerabilities of Cross-chain Bridge Mechanisms * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * --- **Reviews are arranged chronologically from the most recent to the oldest** --- * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * ## Next Steps For Paper 1. - [ ] All Refences / Figures / Diagrams should be genrated from Mendeley / GitHub a. Integrate References with Mendeley b. Remove Duplicated from the Mendeley c. Integrate Figures with GItHub --- 2. - [ ] Integrate Diagram showing funcional components of the bridge Interacting a. Demonstrate the basic funcality of bridges in operational diagram b. In this diagram, show vulnerability of components / and attack vulnerability / possible defence c. Let components and attacks / categorizations be synonymous with the text in the paper d. Develop another Diagram showing defence mechanisms e. Develop a diagram for taxonomy section --- 3. - [ ] Improve Analysis Section a. Make / Efffectiveness of Security Audits / more empirical - collect more data on secuirty audits -- follow pattern in Liyi's paper b. Develop / Emergency Pause / sub-section of the paper --- 4. - [ ] Develop Defence Section a. For analysis parts, Look into effective bridge defence mechanisms from real-life events [ incidents where the protocols defence mechanisms prevented an attack ] b. Develop Money Tracing c. Look into other sub-section such as Bytecode Similarity Analysis / Rescue and Incident Time Frame d. Investigate Defence Mechanisms from other Academic Papers --- 5. - [ ] Improve Paper Structure & Scope a. Integrate Validation Mechanims into Bridge Components b. Explain in Detail Scope of Paper -- Cross-chain only or Multi-chain -- use general taxonomy MTLC / side chain / relay etc which of these is the paper focused on c. -- --- 6. - [ ] Improve Tables a. Make the style of tables synonymous b. Programmatically Colour Tables based on the green and red schemes -- use these links -- https://github.com/xujiahuayz/contango/blob/164b4298bc57ea1a2132c656b5b81208c5540c60/scripts/tabulate_corr.py#L20 -- -- https://github.com/xujiahuayz/contango/blob/main/tables/corr_ETH.tex --- # All Reviews ACM AsiaCCS 2024 Round 2 Paper #861 Reviews and Comments =========================================================================== Paper #861 SoK: Dissecting Attacks and Vulnerabilities of Cross-chain Bridge Mechanisms Review #861A =========================================================================== Overall merit ------------- 1. Reject Reviewer expertise ------------------ 3. Knowledgeable Paper summary ------------- This paper presents an in-depth analysis of the cross-chain bridge mechanism, and summarizes security vulnerabilities in cross-chain bridge protocols. The authors also investigate real-world attacks on cross-chain bridge protocols in recent years and #### Strengths + comprehensive analysis of cross-chain bridge mechanism + a taxonomy of cross-chain bridge protocols #### Weaknesses - the proposed methodology to analyze cross-chain vulnerability lacks new contribution - it is unclear how the proposed framework and system model helps analyze vulnerabilities Comments for authors -------------------- Since the title of this SoK paper indicates that the main goal of the research is to study attacks and vulnerabilities in cross-chain bridge ecosystem, I expect the content could contribute an in-depth analysis against such topics. Unfortunately, the paper emphasizes the knowledge of cross-chain implementations rather than their security issues. Actually, almost half of the paper (from page 1 to page 8, Section 2 to Section 5) is about how cross-chain bridges are implemented and how should we classify them. By comparison, the security analysis part (Section 6) only present a very general and vague procedure and the results (Section 7) are less comprehensive. From Section 6 and Section 7, I cannot understand how the previously introduced knowledge is utilized. And in Section 8 it seems all the investigated attacks are manually classified without providing some insights (e.g., the root causes, the lessons learned), and Table 2 & 3 are very difficult to follow. Therefore, I hardly believe that readers can learn security related knowledge from these parts. In all, an SoK paper should not simply enumerate existing events or just give a classification, but should try to find the in-depth connections between the investigated targets and report some new insights. I do not think this paper achieves this goal and it requires a careful revision to improve its scientific contribution. Review #861B =========================================================================== Overall merit ------------- 1. Reject Reviewer expertise ------------------ 3. Knowledgeable Paper summary ------------- This paper studies the security of cross-chain bridges, which have been popular attack targets in recent years. The paper introduces a taxonomy on cross-chain bridges, and studies recent bridge attacks in real-world. Comments for authors -------------------- I think security of cross-chain bridges are important. However, I am a bit skeptical on this SoK paper. ### What is the difference of cross-chain bridge attacks compared to traditional smart contract attacks? The first question, and probably the most important question, is that, what are the major differences of bridge attacks, compared to traditional smart contract attacks. IMHO, most of the bridge attacks are either due to smart contract bugs, or due to the stupid mistakes by the bridge developers (e.g., lack of access control). I failed to see the major difference of bridge attacks. So I am not quite sure if the paper would have enough academic value. ### Lack of detailed attack cases This paper presents aggregated attack data in Sec 7, and attack categories in Sec 8. While they are both useful, I would appreciate more in-depth case studies on the attacks. For example, the authors can include (simplified) source code of the real-world bridge attacks, and explain the root cause of the attacks. ### What are the takeaways? As an SoK paper, one would expect that it includes important takeaways for the readers to learn. For example, what are the root causes of such attacks? What are the current mitigations? What more can be done to secure the bridges? Such takeaways would help the readers better grasp the key insights of this paper. Right now, the paper does not provide such insights. ## USENIX Security '24 Summer Submissions Review #2125A =========================================================================== Paper summary ------------- In the paper, the authors summarize and provide detailed explanations of cross-chain bridges. They define the structures of cross-chain bridges and which components are commonly used. Furthermore, they analyze cross-chain bridge attacks and statistically analyze the attacked targets and the time of the exploits. Their key insight is that most attacks focus on the source chain contract. Detailed comments for authors ----------------------------- In the interoperability trilemma, the systems extensibility, generality, and trustlessness interfer with each other. After reading the paper, I don't understand why that should be a trilemma. A system's extensibility does not interfere with trustlessness at all. In a light client bridge, one can simply deploy the router contract on the target chains. In a sidechain approach, one can simply start a new node that subscribes to another chain. Both are little to no effort but different levels of trustlessness. Yet, they still make the bridge extensible. In the bridging trilemma, the paper states that native tokens hinder unified liquidity and instant guaranteed finality. While that may be true for unified liquidity, it certainly is not for the finality. The finality requirement is neither dependent on the native token of a blockchain nor dependent on the unified liquidity. To me, the paper does not properly explain why this should be a trilemma and not a dilemma and a general finality constraint. The instant guaranteed finality is part of the inherent structure of the consensus of the underlying blockchain. Ensuring finality has to be done no matter what token or data is bridged. Section 3.3 is a comparison of three different bridges. However, there is actually no comparison in this section. There are three bridges introduced but not compared. Furthermore, all the information about trilemmas before is forgotten about as for none of the bridges the tradeoffs are shown or even mentioned. Table 2 is never explained. It would be interesting to know what "False Top-up" means or an explanation about what these attacks are, i.e., Refund Logic Exploit, ChainID vulnerability exploit, or Rush Attack. As there are exploits shown that are never explained, the paper is not comprehensive. Also, private key compromise is shown in the table, which is a very important type of attack, but never mentioned in the text. There is already a SoK paper (Not Quite Water under the Bridge: Review of Cross-Chain Bridge Hacks) which does a great job explaining bridges, and although not comprehensive either, covers all common attacks. Both papers contain many bugs, for instance the Unrestricted Deposit Emitting (Inconsistent Deposit Logic) are present in both papers. The Verfication Circumventing (Bypassing Signature Verification) is present in both papers as well. Inconsistent Event Parsing (Forwarding Invalid Messages) is the same bug as well. The big difference between the two papers is that Not Quite Water under the Bridge also describes token interface attacks, whereas the Dissecting Attacks and Vulnerabilities paper introduces more attacks on the infrastructure (see 7.4). Apart from this, the Not Quite Water under the Bridge paper better describes the actual bridging process as well as gives for each attack an idea about how to defend against the attack. Therefore, I am not sure about the novelty of the paper in comparison to Not Quite Water under the Bridge. # Minor In 7.2.2, the distributed private key control scheme is mentioned, however, the scheme is never explained. This, together with the hash-timelocks, which is also mentioned but never explained, shows an incomprehensive description of how bridges operate. Framework might be the wrong word as there is no implementation done for a cross-chain bridge, but a conceptualization. There are a lot of sources in the bibliography that do not have a matching URL or a link or publisher. See 7,8,9,13,14,22,51,52,53. Also, there are duplicate entries. # Sidenotes In 4.2.2. there is a missing reference in the paragraph about monetary goals. The last sentence in 5.2 is missing a word. In 6.2. the first sentence is missing a point. Ethics consideration -------------------- 1. No Required changes ---------------- - Describe Table 2. - Explain the bridging process as there is no description of how assets or data are bridged. - Explain why instant guaranteed finality hinders unified liquidity and native tokens. - Actually compare the three bridges in the comparison and classify the bridges according to your systematization. - Explain all bugs shown in Table 2. Reasons to accept the paper --------------------------- - The paper describes all components of common cross-chain bridges. - The classification and systematization of the cross-chain bridges are done well. - The paper describes a well-defined threat model. - The paper gives a great overview over the amount of incidents in the past and at which part of a cross-chain bridge how much money was lost. Reasons to not accept the paper ------------------------------- - The paper is not comprehensive, as there is no description of how assets and data are actually transferred. - A lot of bugs in Table 2 are never explained like Private Key Compromised, Refund Logic Exploit, or the Multi-signature permission vulnerability. - The attacks described are not comprehensive as there are still attacks mentioned but not covered. - There is already a SoK Paper by Sung-Shine Lee et al. "SoK: Not Quite Water Under the Bridge: Review of Cross-Chain Bridge Hacks" that covers many of the vulnerabilities and attacks and even proposes mitigations for the vulnerabilities, therefore, the contribution of the paper is weak. - Schemes like distributed private key controlor Hash Timelocks are mentioned, but never introduced. - The "framework" is a system model that does not define a framework, but rather reeducates about what components are needed in a bridge. Recommended decision -------------------- 4. Reject Writing quality --------------- 3. Adequate Confidence in recommended decision ---------------------------------- 3. Highly confident (would try to convince others) * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * Review #2125B =========================================================================== Paper summary ------------- This SoK paper presents an overview of blockchain bridges and studies ~30 bridge-related security incidents. The authors first provide an overview of different components and types of bridges. Then, they cover three examples. Finally, they provide an analysis of real-world security incidents. Detailed comments for authors ----------------------------- I was very excited when I saw the title of the paper. I agree with the authors that bridges have become increasingly important components of the blockchain ecosystem. It is also true that we have witnessed a substantial increase in attacks against such bridges. Hence, it is important and timely to study this problem. Unfortunately, my excitements waned fairly quickly. I had hoped for an analysis of the core challenges that bridges face when they have to cross the trust boundaries between different chains. Based on these core challenges, the authors could have identified different solutions that projects tried to solve these issues, together with the trade-offs and advantages. This could have also allowed them to identify gaps in the current designs, pointing to possible future avenues of research. Instead, the first part of the paper is mostly a high level summary of basic components of bridges. The key challenges, as discussed in Section 3.2, are mostly a summary of a few blog posts [15-17, 19]. The authors present the interoperability and bridging dilemmas that are mentioned in these blog posts, but they do not critical question or motivate them. Are these trilemmas real? What are the fundamental reasons? What are the possible solutions that exist today? What can be done in the future? All these points remain unanswered. The second part of the paper covers a number of real-world attacks. Table 1 claims to "show a systemisation of bridge attacks and a gap analysis." However, I don't really see a systematization. In fact, almost all entries in this table have a single incident, the top row lists 3. How does this table generalize and provide higher-level grouping (systematization). I also do not see a gap analysis. What can we learn from this table (and corresponding discussion)? What are the key directions for future work? In Section 4, the authors claim that they introduce a comprehensive system and threat model to classify attacks. This model includes three parts: the source chain, the destination chain, and the off-chain components. However, later in Section 6.2 (Figure 2), it seems that 16.1% of the observed 31 incidents cannot be classified by the proposed model (and is in the "Other" category). I think this should have given the authors pause and make them rethink their model given that it already is not sufficient for the small number of studied attacks. Finally, the editorial quality of the paper needs improvements. There are reference in the text that are missing, there are broken sentences, and the references in the bibliography miss relevant data (including publication venues, publication dates, and URLs for blog posts). Also, the data reported in the paper seems inconsistent. From Section 5.2: "We identify 31 real-world attacks on cross-chain bridge protocols .. over the period of 2 years from May 31, 2021 to Sep 30, 2023." From Section 6.3: "Between May 31, 2021, and September 30, 2023, a total of 28 cross-chain bridges experienced security incidents." Ethics consideration -------------------- 1. No Reasons to accept the paper --------------------------- + Blockchain bridges are important, and we have recently seen an increasing number of security incidents that involve bridges Reasons to not accept the paper ------------------------------- - Most of the paper is a literature survey - The analysis of the security incidents is fairly shallow - The editorial quality of the paper is somewhat low Recommended decision -------------------- 4. Reject Writing quality --------------- 4. Needs improvement Confidence in recommended decision ---------------------------------- 3. Highly confident (would try to convince others) --- ## Paper Structure **0 / ABSTRACT** **A / INTRODUCTION** **B / BRIDGE PRELIMINARIES** actors / assets / components **C / BRIDGE MECHANISMS FORMALISATION** mechanism / **D / PROTOCOL COMPARISON** design dynamics / taxonomy / comparison **E / BRIDGE REFERENCE FRAMEWORK** system model / threat model **F / DATA** - In-text academic papers / audit reports / real-world incidents - Sources of Incidents: DeFi Lama Bridge Attacks & Literature **G / ATTACK ANALYSIS** - In-text attack frequency / bridge component targeted / bridge protocol types / security audit - Diagrams attack frequency diagram 1 / attack frequency diagram 2 / bridge component targeted diagram **H/ DEFENCE ANALYSIS** rescue and incident timeframe / bridge component targeted / bridge protocol types / security audit - I / DISCUSSION - J / RELATED WORKS - K / CONCLUSION ## Tasks #### Ali Completed: 1a, 1b, 5a, 2e, Define a bridge and describe an overall bridge mechanism Read on papers which have a 4/5-based taxonomy -> HTLC, relays, side-chain, etc. Include a sub-section of taxonomy based on this overall mechanisms which is applicable to all bridge types you include Have a justification for all example types included Have a justiciation for all example types not included #### Kelly #### Yimika

Syntax	Example	Reference
# Header	Header	基本排版
- Unordered List	Unordered List
1. Ordered List	Ordered List
- [ ] Todo List	Todo List
> Blockquote	Blockquote
Bold font	Bold font
Italics font	Italics font
~~Strikethrough~~	~~Strikethrough~~
19^th^	19^th
H~2~O	H₂O
++Inserted text++	Inserted text
==Marked text==	Marked text
[link text](https:// "title")	Link
![image alt](https:// "title")	Image
`Code`	`Code`	在筆記中貼入程式碼
```javascript var i = 0; ```	`var i = 0;`
:smile:		Emoji list
{%youtube youtube_id %}	Externals
$L^aT_eX$	L^aT_eX
:::info This is a alert area. :::	This is a alert area.