Walrus
Check out Frimz, an AI thinking partner built with Walrus Memory that remembers your past decisions, rejected ideas, and evolving context across sessions, by @tissuesurgeon. There's still time to get your submissions in for Walrus Sessions 8: Chatbots That Remember.
中文: 看看Frimz，一个由Walrus Memory打造的人工智能思维合作伙伴，通过@tissuesurgeon回忆你过去的决策、被拒绝的想法，以及跨会话的演变背景。 还有时间提交您的海象课程8:记住聊天机器人。
#351 🔗 原推
2026-10-05 04:32:29
Walrus
RT @0xmht: This week I’m in Singapore for Sui Basecamp and Token2049 with team @conso_xyz If you’re attending, let’s connect. Excited to meet everyone @SuiNetwork and @WalrusProtocol pushing the ecosystem forward. See you there 🤝 gConso 🐻‍❄️
中文: RT @0xmht:本周我来新加坡参加Sui Basecamp和Token2049，团队团队支持@conso_xyz 如果你去参加，那就联系一下吧。很高兴与所有人见面，@SuiNetwork 和 @WalrusProtocol 推动生态系统发展。 在那里见 🤝 gConso 🐻 ❄️
#350 🔗 原推
2026-10-05 04:23:43
Walrus
RT @ZombMobz: @WalrusProtocol @matterhornso @chain_ofthought @kimblgn @StatescuRazvan @CarrySui @blockticity @ConorBronsdon @sang_wen @chainguard_dev @genspark_ai Very nice! Keep up the expansion. AI needs that memory. Can’t emphasize this enough! Valuable digital assets need to be stored in a secured decentralized location! People will realize this 🔥
#349 🔗 原推
2026-10-05 04:03:45
Walrus
RT @matterhornso: We're very excited to build with Walrus! Persistent memory is a key element for us because it turns AI from an assistant into a true coworker... One that understands context, remembers what came before, and can pick up where our users left off. Particularly relevant in Web3. Great work all, and looking forward to what's ahead!
#348 🔗 原推
2026-10-05 04:03:11
Walrus
RT @0xd34th: The basecamp venue is massive. Mind-blowing announcements on the 7th and 8th. Livestreamed as well. Sui you there 💧 https://twitter.com/0xd34th/status/2106726100990640138/photo/1
media 0 共 1 项
#347 🔗 原推
2026-10-05 04:02:46
Walrus
RT @EmanAbio: Big big week. LFG 🔥
中文: RT @EmanAbio:大大周。LFG 🔥
#346 🔗 原推
2026-10-05 03:51:52
Walrus
RT @SuiNetwork: 💎 Diamond: - @EVE_Frontier: Hardcore space survival game with a player-run economy, creators of EVE Online - @WalrusProtocol: The data platform built for the demands of AI
中文: RT @SuiNetwork:💎 钻石: - @EVE_Frontier:玩家经营经济型硬核太空生存游戏，EVE Online 的创作者 - @WalrusProtocol:为人工智能需求而构建的数据平台
#345 🔗 原推
2026-10-05 03:50:46
Walrus
RT @SuiNetwork: Sui Basecamp 2026 is three days out. You don’t build the rails of the agentic economy alone. Meet the 12 teams powering Sui Basecamp 2026. All to answer one question: what does finance look like when most of it isn’t human? Oct 7-8, Marina Bay Sands, Singapore. https://twitter.com/SuiNetwork/status/2106821414477287900/photo/1
中文: RT @SuiNetwork:苏·巴斯坎普2026号即将推出三天。 你无法独自建立经济的轨道。认识一下为苏·巴斯坎普2026号提供动力的12支队伍。 回答一个问题:当大多数金融并非人类时，它会是什么样子? 10月7日至8日，新加坡滨海湾金沙。
media 0 共 1 项
#344 🔗 原推
2026-10-05 03:45:09
Walrus
Models change. Agent frameworks change. Tools change. Your data has to follow. On October 7, we're giving you a new way to keep your data useful, portable, and under your control. https://twitter.com/WalrusProtocol/status/2106079420293530035/photo/1
media 0 共 1 项
#343 🔗 原推
2026-10-02 17:50:37
Walrus
If you are attending Sui Basecamp in Singapore next week, find the Walrus team on the main floor: Keynotes: Mainstage presentations from @kostascrypto, @kimblgn, and @EmanAbio Plus, catch @DLougaris on the AI Builder Lab stage. Partner Showcases: Real-world use cases with @astros_ag , @AlliumLabs , and @matterhornso Stop by our exhibition booth and say hi.
中文: 如果你下周去新加坡参加隋大本营，请在主楼层找到海象队: 主题演讲:来自 @kostascrypto、@kimblgn 和 @EmanAbio 的主舞台演示 此外，在AI Builder实验室舞台上关注@DLougaris。 合作伙伴展示:使用 @astros_ag、@AlliumLabs 和 @matterhornso 的真实使用案例 到我们的展台去，说个好话。
media 0 共 1 项
#342 🔗 原推
2026-10-02 14:59:15
Walrus
If you are attending Sui Basecamp in Singapore next week, find the Walrus team on the main floor: Keynotes: Mainstage presentations from @kostascrypto, @kimblgn, and @EmanAbio Plus, catch @DLougaris on the AI Builder Lab stage. Partner Showcases: Real-world use cases with @astros_ag , @AlliumLabs , and @matterhornso Stop by our exhibition booth and say hi.
中文: 如果你下周去新加坡参加隋大本营，请在主楼层找到海象队: 主题演讲:来自 @kostascrypto、@kimblgn 和 @EmanAbio 的主舞台演示 此外，在AI Builder实验室舞台上关注@DLougaris。 合作伙伴展示:使用 @astros_ag、@AlliumLabs 和 @matterhornso 的真实使用案例 到我们的展台去，说个好话。
media 0 共 1 项
#341 🔗 原推
2026-10-02 14:52:37
Walrus
If you're attending Sui Basecamp in Singapore next week, find the Walrus team on the main floor: ✔️ Keynotes: Mainstage presentations from @kostascrypto, @kimblgn, and @EmanAbio ✔️ Partner Showcases: Real-world use cases with @astros_ag , @AlliumLabs, and @matterhornso Stop by the booth to see persistent state running in real time.
中文: 下周前往新加坡的隋大本营，请在主楼层找到海象队: ✔️ 主题演讲:来自 @kostascrypto、@kimblgn 和 @EmanAbio 的主舞台演示 ✔️ 合作伙伴展示:使用 @astros_ag、@AlliumLabs 和 @matterhornso 的真实使用案例 在展位旁停下来，实时查看持续的状态。
media 0 共 1 项
#340 🔗 原推
2026-10-02 09:57:00
Walrus
"If you rent a house... you build no long-term equity." @thomsonreuters CTO @JoelHron on @chain_ofthought: why renting AI models builds zero equity, and why owning your own model lets your team's expertise compound over time. Full episode presented by Walrus: https://youtu.be/CQpAVPSYPxI
🎬
视频
#339 🔗 原推
2026-10-01 22:17:00
Walrus
@kimblgn recently sat down with the Lessons in Product Management podcast to discuss product leadership, building for an agentic future, and what happens when the end user is no longer a human, but an AI agent. A few key takeaways: • How product leadership evolves when agents become the primary consumers of your stack • Designing infrastructure with provenance, auditability, and user control at the core • Why durable, portable state matters when workflows span across models and runtimes Listen to the full episode: https://open.spotify.com/episode/4wJabCwDVGX52mAZRN74ia
中文: @kimblgn 最近与“产品管理课程”播客进行了讨论，讨论产品领导力、构建一个具有代理性的未来，以及当最终用户不再是人类，而是人工智能代理时会发生什么。 几个关键要点: • 当代理成为您堆栈的主要消费者时，产品领导地位如何演变 • 以出处、可审计性和用户控制为核心设计基础设施 • 当工作流程跨模型和运行时时，为何耐用且便携的状态至关重要 收听完整节目:
#338 🔗 原推
2026-10-01 19:45:15
Walrus
RT @ConorBronsdon: Genspark co-founder @sang_wen: "Salesforce went headless. We're building the head." Back in September, Wen (co-founder & COO of @genspark_ai) and I got into: - why frontier labs build engines and Genspark builds the car - grading the deliverable, not the model's intelligence - why a meeting note is where the work starts, not where it ends Disclosure: Genspark gave me a SecondBrain Note review unit at no cost. They did not sponsor this episode. Brought to you by: • @WalrusProtocol - Walrus Memory gives AI agents portable, verifiable memory that carries context across apps, sessions, and other agents: https://walrus.xyz/cot • @SvixHQ - reliable webhooks for startups and the Fortune 500. Qualified startups get $12,000 in credits, YC companies $50,000: https://link.svix.com/cot • @inngest - durable execution for agents in production. Failed steps retry; completed steps are saved and skipped: https://inngest.link/cot-x • @g2i_ai - a decade of vetting engineers, now turned on reviewing the RL environments, evals, and training data models learn from: https://fandf.co/3SFxVm6 Look up the @chain_ofthought podcast on YouTube, Spotify, Apple Podcasts (links below), or wherever you listen. Thanks, Wen!
中文: RT @ConorBronsdon:Gensparg联合创始人@sang_wen:“Salesforce 无头无电。我们正在建立头脑。 早在9月份，我和温(@genspark_ai的联合创始人兼首席运营官)就进入了: - 前沿实验室为何制造发动机，而Gensparks制造汽车 - 对可交付性进行分级，而不是对模型的智能进行评分 - 会议说明为何是工作开始的起点，而不是终点 免责声明:Gensparks 免费为我提供了 SecosenBrain Note 评测单元。他们没有赞助本集。 由以下人带给您: • @WalrusProtocol - Walrus Memory 为AI代理提供可移植且可验证的存储，可在应用程序、会话和其他代理中存储上下文: • @SvixHQ——适用于初创企业和财富500强的可靠网络钩子。合格的初创企业获得12000美元的积分，YC公司获得5万美元: • @inngest——生产中对代理产品的持久执行。步骤重试失败;已完成的步骤被保存并跳过: • @g2i_ai——十年来，工程师们经过了十年的审查，如今开始重新审视RL的环境、评估结果以及从以下内容中学习的数据模型: 在 YouTube、Spotify、Apple 播客(下方链接)或您收听的任何地方查看 @chain_ofthought 播客。谢谢，温!
🎬
视频
#337 🔗 原推
2026-10-01 16:33:46
Walrus
RT @SabrinaMeissnr: @WalrusProtocol @matterhornso I recently tested @matterhornso and is a really strong product. Bullish!
中文: RT @SabrinaMeissnr:@WalrusProtocol @matterhornso 我最近测试过 @matterhornso，是一款非常强大的产品。欺负人!
#336 🔗 原推
2026-10-01 14:58:01
Walrus
RT @EdCriptoFi: Most chatbots forget you the second you close the tab. WalCoach is a coaching companion with 7 mentors that share one memory of you. That memory is not in my database: it lives on @WalrusProtocol, in an account you own. 2 minutes, real session @SuiDevelopers #WalrusSessions 🔗 https://walcoach.vercel.app/
中文: RT @EdCriptoFi:大多数聊天机器人在关闭标签页后会忘记你。 沃尔科奇是一位教练伙伴，拥有7位导师，共同拥有你的一段记忆。 该内存不在我的数据库中:它位于 @WalrusProtocol 上，位于您拥有的账户中。 2分钟，真实会话 @SuiDevelopers #华勒斯赛人 🔗
🎬
视频
#335 🔗 原推
2026-10-01 14:54:14
Walrus
RT @EdCriptoFi: WalNotes + WalCoach All in one place. @wal_notes
中文: RT @EdCriptoFi:WalNotes + WalCoach 一切在一个地方。 @wal_notes
#334 🔗 原推
2026-10-01 14:54:06
Walrus
Expanding context windows handles temporary scale, but true agent autonomy requires persistent state across application restarts. At Sui Basecamp next week, we’re showing builders what happens when you give AI memory a clean UI. Where do context limits start breaking your current stack?
中文: 扩展上下文窗口可处理临时规模，但真正的代理自主权要求在应用程序重启时持续状态。 下周在Sui Basecamp，我们将向构建者展示当你为AI内存提供一个清晰的用户界面时会发生什么。 上下文限制从何处开始打破你当前的堆栈?
media 0 共 1 项
#333 🔗 原推
2026-10-01 10:29:00
Walrus
Subscribe to the monthly newsletter for full updates: https://walrus.xyz/newsletter/
中文: 订阅月度简报以获取完整更新:
#332 🔗 原推
2026-09-30 23:20:46
Walrus
September was a busy month at Walrus. Here's everything that shipped: @matterhornso will make Walrus Memory its default backend. Project context, deployment history, audit results, and agent conversations now persist across sessions and stay shared across team members. Plus, 5,000+ @ASI_Alliance users get decentralized AI models backed by Walrus. • $7.7B in Real-World Assets: @Blockticity is bringing 1M+ authenticated trade records to Walrus. • Sui Overflow Winners: Kraterion built by @StatescuRazvan (S3-compatible sealed files), Suize (one-call verified sites), and Carry (verifiable memory receipts). • Walrus Session 8 is LIVE: Build chatbots that remember, now through October 9. • Walrus is a presenting sponsor of season four of @chain_ofthought. Catch the latest episodes with @genspark_ai’s @sang_wen on building a second brain for agents, and @chainguard_dev CEO @dlorenc on why time to exploit has gone negative. Find us in Singapore at Sui Basecamp (Oct 7–8). Stop by the Walrus booth or catch our Head of Product @kimblgn, on stage.
media 0 共 1 项
#331 🔗 原推
2026-09-30 23:19:30
Walrus
RT @lexlex99722746: @WalrusProtocol Session 8 has kicked off. Check it out: https://www.deepsurge.xyz/hackathons/c0141a4a-21be-4009-bc63-7c168608c849 The are running a promo: https://walform.wal.app/f?formId=0x38a736485349b133604c1caf286d669b4b774d16f0a26120ce839cad245baeef #WalrusMemory https://twitter.com/lexlex99722746/status/2105417295497474057/photo/1
media 0 共 1 项
#330 🔗 原推
2026-09-30 23:11:07
Walrus
RT @AKS_2683: @conso_xyz @WalrusProtocol And Conso is following that path.
中文: RT @AKS_2683:@conso_xyz @WalrusProtocol 和 Conso 正在沿着这条路走下去。
#329 🔗 原推
2026-09-30 18:43:14
Walrus
RT @conso_xyz: @WalrusProtocol Context is what makes agents powerful WALRUS is building the infrastructure to power it
中文: RT @conso_xyz:@WalrusProtocol 上下文正是使代理功能强大的原因 WALRUS 正在建设其供电基础设施
#328 🔗 原推
2026-09-30 18:43:13
Walrus
From shared agent memory on Matterhorn to $7.7B in real-world assets, here’s everything that shipped across the ecosystem this month: @matterhornso x Walrus: Matterhorn will make Walrus Memory its default backend. Project context, deployment history, audit results, and agent conversations now persist across sessions and stay shared across team members. Plus, 5,000+ @ASI_Alliance users get decentralized AI models backed by Walrus. • $7.7B in Real-World Assets: @Blockticity is bringing 1M+ authenticated trade records to Walrus. • Sui Overflow Winners: Kraterion by @StatescuRazvan (S3-compatible sealed files), Suize (one-call verified sites), and @CarrySui (verifiable memory receipts). • Walrus Session 8 is Live: Build chatbots that remember, now through October 9. • Walrus Memory is a presenting sponsor of season four of @chain_ofthought. Catch the latest episodes with @genspark_ai’s @sang_wen on building a second brain for agents, and Chainguard CEO @dlorenc on why time to exploit has gone negative. • Find us in Singapore at Sui Basecamp next week. Stop by the Walrus booth or catch our Head of Product, @KimberlyLogan on stage. Subscribe to the newsletter for full monthly updates: https://walrus.xyz/newsletter/
media 0 共 1 项
#327 🔗 原推
2026-09-30 18:42:31
Walrus
@matterhornso and Walrus are bringing persistent memory to AI-assisted blockchain workflows. Instead of AI sessions resetting and forcing you to re-explain project context from scratch, your work now persists across sessions and teams. Read the announcement: https://blog.walrus.xyz/matterhorn-walrus-persistent-memory-ai-workflows/
中文: @matterhornso 和 Walrus 正在为人工智能辅助的区块链工作流程带来持久内存。 不再让人工智能会话重置并强制你从零开始重新解释项目背景，而是在会议和团队中持续进行工作。 阅读公告:
media 0 共 1 项
#326 🔗 原推
2026-09-30 16:30:31
Walrus
AI agents can act. But your data is what makes them useful. Files, history, context, and everything your apps and agents accumulate becomes part of your edge. In the #AI era, that data needs to move with you. October 7th: A new way to connect your data across apps and AI tools is coming.
中文: 人工智能代理可以采取行动。但你的数据是使它们变得有用的。 文件、历史、上下文以及你的应用和代理所积累的一切都成为你优势的一部分。 在#AI时代，这些数据需要与你同在。 10月7日:一种通过应用程序和人工智能工具连接数据的新方式即将到来。
media 0 共 1 项
#325 🔗 原推
2026-09-30 15:56:00
Walrus
TODAY: We're sitting down with Satyam Bansal to unpack @redsentinel_ai 's build, its integration with Walrus and Sui, and how they approach agent security. Tune in for the technical teardown and an early look at Walrus Session 9.
#324 🔗 原推
2026-09-30 09:56:00
Walrus
RT @redsentinel_ai: For most of us, LLMs are a black box. We type a prompt, we get a response, and that's it. We don't really know what happens behind the scenes: how the model thinks, how it reasons, how it produces an answer one token at a time. Many of us have never even heard of prompt injection or jailbreaking. And with so many models out there, how do we pick the one that fits our needs, serves us best, and, most importantly, stays secure and behaves the way we expect? Red Sentinel is an AI arena where you can learn all of this by doing, and where your skills turn into real economic rewards. The loop is simple: build an AI defense, put money behind it, and let the internet test it. Every failed attack grows the reward pool, and a verified breach releases the full bounty to the attacker.
中文: RT @redsentinel_ai:对我们大多数人来说，LLM 是一个黑盒子。我们输入提示，得到回复，就这么了。我们并不真正了解幕后发生了什么:模型如何思考、如何推理，以及它如何一次生成一个代币的答案。我们中的许多人甚至从未听说过及时注射或越狱。有这么多模型，我们如何选择符合我们需求、最符合我们需求、最符合我们需求的模式，以及最重要的是，保持安全并按照我们期望的方式行事? 红色哨兵是一个人工智能领域，你可以通过这样做来学习这一切，以及你的技能转化为真正的经济回报。 循环很简单:建立人工智能防御，投入资金支持，并让互联网进行测试。每一次失败的攻击都会增加奖励池，而经过验证的泄露行为会向攻击者释放全部赏金。
#323 🔗 原推
2026-09-30 04:32:38
Walrus
RT @williamm168: 💬 1,000 chat messages shouldn't mean 1,000 onchain transactions. That's the design choice behind Sui Stack Messaging SDK Beta, live on @SuiNetwork and @WalrusProtocol mainnet. Here's the split: → Sui records who belongs to a channel and what they're allowed to do. → A relayer delivers messages in real time. It checks membership but never sees plaintext. → Seal encrypts messages before they leave your device. → Walrus stores encrypted backups so you can restore history on another device. Sui decides who's in the room. It doesn't have to carry every sentence. That could make wallet-linked guild chats and in-app support feel like chat, not a series of transactions. This is an SDK for builders, not a finished messenger you can download today. Which would you use first: a private guild chat or support inside a Sui app?
中文: RT @williamm168:1000条聊天消息不应指1000次链上交易。 这是 Sui Stack 消息 SDK 测试版的设计选择，它在 @SuiNetwork 和 @WalrusProtocol 主网上进行直播。 以下是分拆: → Sui 记录属于频道的人以及他们被允许做什么。 → 中继器实时传递消息。它检查会员资格，但从未看到明文。 → Seal 会在消息离开您的设备之前对其进行加密。 → 海象存储加密备份，以便您能够在其他设备上恢复历史记录。 苏决定房间里有谁。不必承载每一句话。 这可能让钱包关联的公会聊天和应用内支持感觉像是聊天，而不是一系列交易。 这是一个面向构建者的SDK，而不是您今天可以下载的成品信使。 你首先会使用哪种:在Sui应用程序内进行私人公会聊天或支持?
media 0 共 1 项
#322 🔗 原推
2026-09-30 04:31:55
Walrus
1,000 chat messages shouldn't mean 1,000 onchain transactions. Great Session 8 build by @williamm168 leveraging the Sui Stack Messaging SDK and Walrus to keep chat history persistent and encrypted without friction.
中文: 1000条聊天消息不应意味着1000次链上交易。 由 @williamm168 构建的 Great Sessep 8，利用 Sui Stack 消息SDK和Walrus，在不受摩擦的情况下保持聊天记录的持久性和加密。
#321 🔗 原推
2026-09-30 04:31:51
Walrus
Join us TOMORROW on Discord for an AMA with Satyam Bansal from @redsentinel_ai. We’ll break down how Red Sentinel builds AI security on Walrus and @SuiNetwork, how their architecture works, and why verifiable context matters. Plus, a preview of what’s coming in the next Walrus Session. Date: September 30, 2026 Time: 12:00 PM UTC Location: Walrus Discord (https://t.co/blpJ6cyRIz)
media 0 共 1 项
#320 🔗 原推
2026-09-29 19:40:00
Walrus
RT @kimblgn: I’ve spent close to 20 years building products, and one of the things I enjoy most is making powerful technology useful to more people That’s why I’m so thrilled about what we’ll be unveiling at Sui Basecamp for @WalrusProtocol I think it’s going to change how people experience Walrus
中文: RT @kimblgn:我花了近20年时间制作产品，其中最喜欢的一件事就是让强大的技术对更多人有用 这就是为什么我对我们将在苏大本营为@WalrusProtocol发布的内容感到非常兴奋 我认为这将改变人们体验海象的方式
#319 🔗 原推
2026-09-29 18:48:25
Walrus
Models get smarter, but context disappears the moment a session ends. Ahead of Sui Basecamp, we're documenting where stateful AI infrastructure needs to go. What is the single biggest workflow failure AI agents cause by 2027 if persistent memory stays unsolved? Let us know in the replies.
中文: 模型变得更智能，但会话结束时，上下文就会消失。 在苏大本营之前，我们正在记录有状态的人工智能基础设施需要走向何方。 到2027年，如果持久性内存仍未解决，人工智能代理导致的单项最大工作流故障是什么? 请在回复中告诉我们。
media 0 共 1 项
#318 🔗 原推
2026-09-29 17:50:29
Walrus
RT @ConorBronsdon: @WalrusProtocol @chainguard_dev @lorenc_dan @chain_ofthought Thank you for sponsoring @chain_ofthought - and thank you @lorenc_dan for coming on the podcast!
中文: RT @ConorBronsdon:@WalrusProtocol @chaingguard_dev @lorenc_dan @chain_ofthought 感谢您赞助 @chain_ofthought——并感谢@lorenc_dan 上播客!
#317 🔗 原推
2026-09-28 23:52:31
Walrus
RT @Konoput: What happens when the chatbot does not own the memory? Walrus Session 8 is Chatbots That Remember, open through Oct 9, $2,000 WAL on the official pool. @WalrusProtocol $WAL #Walrus https://twitter.com/Konoput/status/2104688593742786682/photo/1
media 0 共 1 项
#316 🔗 原推
2026-09-28 23:52:10
Walrus
Open models lowered the cost of generating attacks, shifting the bottleneck from finding bugs to fixing them. @chainguard_dev CEO @lorenc_dan joined @ConorBronsdon on @chain_ofthought to break down why defenders need better speed. Presented by Walrus: https://twitter.com/WalrusProtocol/status/2104652452918333516/video/1
中文: 开放模型降低了生成攻击的成本，将瓶颈从发现漏洞转向了修复。 @chaingguard_dev 首席执行官 @lorenc_dan 与 @ConorBronsdon 一起在 @chain_Ofthought 上思考，以说明防守球员需要更快速度的原因。 由 Walrus 呈现:
🎬
视频
#315 🔗 原推
2026-09-28 19:20:21
Walrus
RT @eazitechh: @kimblgn @WalrusProtocol Been using them with my agent, they make my workflow easier A must have for every developer building with Walrus
中文: RT @eazitechh:@kimblgn @WalrusProtocol 一直与我的经纪人一起使用，他们让我的工作流程更加轻松 每座使用 Walrus 的开发者都必须拥有
#314 🔗 原推
2026-09-28 15:15:55
Walrus
RT @tehMoonwalkeR: "I want to choose whichever tool works best for me and bring my saved context along. That’s what we’re solving with @WalrusProtocol Memory, and there’s a lot more we want to do here" This is the future of AI
中文: RT @tehMoonwalkeR:“我想选择最适合我的工具，并为我保存的上下文带来支持。这就是我们使用 @WalrusProtocol 内存来解决的问题，而我们在这里还有很多想要做的事情。 这就是人工智能的未来
#313 🔗 原推
2026-09-28 15:15:49
Walrus
RT @tetreum: @WalrusProtocol portable memory is the real unlock 🧠
中文: RT @tetreum:@WalrusProtocol 便携式内存才是真正的解锁 🧠
#312 🔗 原推
2026-09-28 15:15:22
Walrus
RT @kimblgn: If you’re building with @WalrusProtocol, take a look at these agent skills They cover everything from building and publishing Walrus Sites to integrating Walrus storage into your applications Put them to work and tell us where we can improve https://twitter.com/kimblgn/status/2104297397103014270/photo/1
中文: RT @kimblgn:如果你使用 @WalrusProtocol 进行构建，请了解这些代理技能 涵盖从构建和发布 Walrus Sites 到将 Walrus 存储集成到您的应用程序中的方方面面 让他们去工作，告诉我们在哪里可以改进
media 0 共 1 项
#311 🔗 原推
2026-09-27 19:57:59
Walrus
Cross-mentor memory, zero database, and an honest breakdown of the technical friction and trade-offs. 👏 @EdCriptoFi built WalCoach for Walrus Session 8, showing how a single encrypted memory layer on Walrus lets multiple personas share context without losing history across sessions. Walrus Session 8 runs through October 9, still plenty of time to submit your chatbot build.
中文: 跨年内存、零数据库，以及技术摩擦与权衡的真实分解。👏 @EdCriptoFi 为 Walrus Session 8 构建了 WalCoach，展示了 Walrus 上单个加密内存层如何使多个角色在会话中共享上下文而不会丢失历史。 《海象会话8》将持续到10月9日，但仍有充足的时间提交您的聊天机器人版本。
#310 🔗 原推
2026-09-26 11:21:00
Walrus
RT @CarrySui: The key for us was simple: authorization has to happen before memory reaches the model, not after. Thank you @WalrusProtocol for diving into what we are building 😁🙏
中文: RT @CarrySui:对我们来说，关键很简单:在内存到达模型之前必须进行授权，而不是在模型之后。 感谢@WalrusProtocol深入了解我们正在构建的内容😁🙏
#309 🔗 原推
2026-09-25 20:34:38
Walrus
RT @LeeMorgan_X: @WalrusProtocol Sounds like Walrus Memory can greatly enhance AI agent workflows by providing seamless, language compatible persistent memory solutions. Exciting stuff!
中文: RT @LeeMorgan_X:@WalrusProtocol 类似 Walrus Memory 可以通过提供无缝、语言兼容的持久内存解决方案，极大地提升 AI 代理工作流程。令人兴奋的东西!
#308 🔗 原推
2026-09-25 19:14:48
Walrus
RT @LeeMorgan_X: @WalrusProtocol Got it! Thanks for the info on Walrus Memory, which seems to solve AI agent's limited memory issues.
中文: RT @LeeMorgan_X:@WalrusProtocol 得到了它!感谢提供有关 Walrus Memory 的信息，该信息似乎解决了人工智能代理有限的内存问题。
#307 🔗 原推
2026-09-25 19:14:44
Walrus
Audit logs and proofs aren't the same thing. @CarrySui uses Walrus Memory to give agents verifiable answer receipts and pre-retrieval access control, so unapproved memory never enters the model context. Deep dive on the build: https://blog.walrus.xyz/carry-verifiable-agent-answers-built-on-walrus-memory/
中文: 审计日志和证明不是同一件事。 @CarrySui 使用 Walrus Memory 为代理提供可验证的接送凭证和预审访问控制，因此未经批准的内存永远不会进入模型上下文。 深入探索该建筑:
#306 🔗 原推
2026-09-25 19:12:51
Walrus
RT @EpochSui: New on Epoch Agents: a bubble map for any Sui token. Paste a coin type and see who holds it, which wallets are linked by transfers, and how much really circulates once pools, vesting vaults and burns are set apart. Free for humans, computed in your browser straight from the chain: https://agents.epochsui.com/bubbles For agents it is a paid API over x402, one call per map. If the map cannot be built, you are not charged. The app runs on @WalrusProtocol, the payments settle on @SuiNetwork.
中文: RT @EpochSui:Epoch Agents 上的新版:任何 Sui 代币的气泡地图。 粘贴一种硬币类型，看看是谁持有的，哪些钱包通过转账连接，以及一旦泳池、毛库和烧伤物被分开，实际流通量有多大。 适用于人类的免费，直接在浏览器中通过链式电脑进行: 对于代理而言，它是一个超过 x402 的付费 API，每个映射一次调用。如果无法建造地图，则您不会收费。 该应用程序在@WalrusProtocol上运行，付款在@SuiNetwork上结算。
media 0 共 1 项
#305 🔗 原推
2026-09-25 18:46:06
Walrus
We're less than two weeks away until Sui Basecamp kicks off. Make sure you’re following along, October 7 is going to be huge. See you in Singapore 🇸🇬
中文: 距离苏·巴斯坎普开场还不到两周。 一定要关注，10月7日将会非常巨大。在新加坡见 🇸🇬
#304 🔗 原推
2026-09-25 18:44:33
Walrus
RT @ledoraeth: I’m ready for @SuiNetwork Basecamp 💧 Are you ? 🦭 https://twitter.com/ledoraeth/status/2103391107690443072/video/1
中文: RT @ledoraeth:我已准备好参加 @SuiNetwork Basecamp 💧 你是吗?🦭
🎬
视频
#303 🔗 原推
2026-09-25 15:09:08
Walrus
RT @Smigglemedia: It’s always amazing building better with AI Learnt about @Daya_HQ partnership & integrations with @SuiNetwork @suidevelopers -Stay positioned for opportunities on @DeepBookonSui @tradeonhudi @WalrusProtocol @CurrentSUI It’s Lockedin mode with @cyberX___ and the gang during @SuiHubAfrica sessions This is your sign to join us at Suihub, learn directly what the founders @EmanAbio @EvanWeb3 @kostascrypto are shipping and contribute to build the best L1 in web3
media 0 共 4 项 media 1 media 2 media 3
#302 🔗 原推
2026-09-24 16:09:20
Walrus
RT @Smigglemedia: @WalrusProtocol @kostascrypto @HumanXCo This looks so cool. Walrus events are awesome🦭💙
#301 🔗 原推
2026-09-23 18:36:45
Walrus
RT @SuiHubAfrica: What if your chatbot could actually remember? Walrus Session 8: Chatbots That Remember is LIVE. Build a chatbot with persistent state, deploy it, and share a before/after write-up showing what changed. $2,000 WAL prize pool 📅 Sep 18 – Oct 9 https://twitter.com/SuiHubAfrica/status/2102790611724120445/photo/1
中文: RT @SuiHubAfrica:如果你的聊天机器人真的能记住呢? 海象会话8:记住的聊天机器人是实时的。 构建一个具有持久状态的聊天机器人，部署它，并分享一个在撰写前/之后显示的更改内容。 2000美元WAL奖金池 📅 9月18日至10月9日
media 0 共 1 项
#300 🔗 原推
2026-09-23 17:29:33
Walrus
RT @Philose: Walrus IRL events are the absolute best 🥺 $WAL connect 🦭
中文: RT @Philose:海象IRL赛事绝对最好🥺 $WAL 连接 🦭
#299 🔗 原推
2026-09-23 17:29:11
Walrus
@kostascrypto and the rest of the Walrus team is on the ground at @HumanXCo in Amsterdam. 🇳🇱 We’re bringing Walrus Memory to the center of the AI stack, solving context loss for multi-agent systems. #HumanX2026 #AI #WalrusMemory https://twitter.com/WalrusProtocol/status/2102775103750848922/photo/1
中文: @kostascrypto 和 Walrus 团队的其他成员将在阿姆斯特丹的 @HumanXCo 上场。🇳🇱 我们将将 Walrus Memory 引入人工智能堆栈，解决多智能系统的上下文损耗问题。 #HumanX2026 #人工智能 #WalrusMemory
media 0 共 2 项 media 1
#298 🔗 原推
2026-09-23 15:00:27
Walrus
One question from the @HumanXCo room: "Is this for personal use or for companies? @kostascrypto's answer: "I believe that the best customers will be companies. You need to be able to have selective access to data based on where you live, what you're using LLMs for, and more. On a personal level though — you share more with your LLMs and your agents that you may even your best friend or your partner. Walrus Memory helps you protect that data, as well as decide when you want to restrict or move it."
中文: 来自@HumanXCo房间的一个问题: 这是供个人使用还是企业使用? @kostascrypto 的回答:“我相信最优秀的客户会是公司。您需要能够根据居住地、使用LLM的用途等数据进行选择性访问。 从个人层面来说——你与你的LLM和经纪人分享更多内容，甚至可能分享你最好的朋友或伴侣。海象记忆有助于保护这些数据，并决定何时需要限制或移动数据。
media 0 共 1 项
#297 🔗 原推
2026-09-23 13:56:42
Walrus
There’s a reason you haven’t switched LLMs: because you’re scared to lose the history and context you’ve shared with it.” Onstage at @HumanXCo Amsterdam, @kostascrypto demos how Walrus Memory makes switching easy, and how to take your data with you. https://twitter.com/WalrusProtocol/status/2102753541777703140/photo/1
中文: 你没有更换LLM的原因有:因为你害怕失去与之分享的历史和背景。 在 @HumanXCo Amsterdam 的 @kostascrypto 上，演示 Walrus Memory 如何轻松切换，以及如何将数据与您一起获取。
media 0 共 1 项
#296 🔗 原推
2026-09-23 13:34:46
Walrus
What happens when your AI agent’s permanent memory accidentally stores a client's real name? For Walrus Session 7, @UyLeQuoc audited Continuum (a cross-tool memory prompt) and fixed two critical privacy and ranking flaws. The fixes: → Local client codenaming so private identities stay off-chain → Monthly namespaces to force time-based sorting over semantic rank Read the full security audit and grab the prompt: https://inkray.xyz/article?id=i-gave-my-ai-a-memory-it-can-never-delete-then-i-read-what-it-was-about-to-write-701fd1d0f587c800
中文: 当人工智能代理的永久内存意外地存储客户真实姓名时会发生什么? 在《海象7》中，@UyLeQuoc 对Continuum进行了审计(一种跨工具内存提示)，并修复了两个关键的隐私和排名缺陷。 修复方法: → 本地客户端代号，以便私密身份保持脱链 → 每月命名空间，以强制基于时间的排序，以克服语义等级 阅读完整的安全审核并获取提示:
media 0 共 1 项
#295 🔗 原推
2026-09-22 21:32:00
Walrus
RT @Philose: @WalrusProtocol Walrus spotlight 🦭
中文: RT @Philose:@WalrusProtocol Walrus 聚光灯 🦭
#294 🔗 原推
2026-09-22 17:35:44
Walrus
Most agent memory is a black box. If an agent recalls context, can you prove where it came from or whether access was authorized? Build Notes #3 looks at Carry, a Sui Overflow winner bringing verifiable answer receipts and pre-retrieval access control to Walrus Memory. Technical breakdown on the blog: https://blog.walrus.xyz/carry-verifiable-agent-answers-built-on-walrus-memory/
#293 🔗 原推
2026-09-22 17:12:10
Walrus
RT @kimblgn: We built Walrus Memory with the expectation that people will keep finding better AI tools. They should be free to use them without rebuilding the context that made their previous tools useful I think preserving that freedom is one of the most important product decisions we can make now
中文: RT @kimblgn:我们开发了 Walrus Memory，期望人们能不断找到更好的人工智能工具。他们应该可以自由地使用它们，而无需重建那些使他们之前工具有用的环境 我认为，维护这种自由是我们现在能够做出的最重要的产品决策之一
#292 🔗 原推
2026-09-22 14:48:00
Walrus
Happening TOMORROW at @HumanXCo. Catch @kostascrypto live at 3:00 PM CEST for an interactive demo on persistence, state, and Walrus Memory. There's still time to add it to your schedule. See you there 🇳🇱
中文: 明天在@HumanXCo上发生。 观看@kostascrypto 直播时间:美国东部时间下午3点，观看关于持久性、状态和海王记忆的互动演示。 还有时间将其添加到你的日程安排中。在那里见 🇳🇱
#291 🔗 原推
2026-09-22 14:46:09
Walrus
RT @SuiNetworkCN: AI Agent 的能力，取决于它能否证明自己所依赖的数据值得信赖。 💎 @WalrusProtocol 正式成为 #SuiBasecamp 钻石级赞助商！ Walrus 是专为满足 AI 需求而打造的数据平台，让数据能够自由迁移、受到保护，并且任何人都可以验证。由打造 Sui 的前 Meta 工程师团队创建。 https://twitter.com/SuiNetworkCN/status/2102245023953916344/photo/1
media 0 共 1 项
#290 🔗 原推
2026-09-22 14:43:31
Walrus
RT @kostascrypto: IMHO the portable, shared & encrypted AI memory angle of Walrus + Sui is one of the best things blockchains ever offered
#289 🔗 原推
2026-09-22 14:43:27
Walrus
RT @kakii120: The future of digital reputation is here. ⚡ @conso_xyz is building a Consumer Reputation Layer powered by @WalrusProtocol & @SuiNetwork. Your activity. Your reputation. Your next opportunity. 🐻‍❄️ https://twitter.com/kakii120/status/2102381538700591310/photo/1
中文: RT @kakii120:数字声誉的未来已到来。⚡ @conso_xyz 正在构建由 @WalrusProtocol & @SuiNetwork 提供支持的消费者声誉图层。 你的活动。你的名声。你的下一个机会。🐻 ❄️
media 0 共 1 项
#288 🔗 原推
2026-09-22 14:41:09
Walrus
RT @LLuciano_BTC: @SuiNetwork @WalrusProtocol AI needs verifiable data. Walrus solves it on Sui.
中文: RT @LLuciano_BTC:@SuiNetwork @WalrusProtocol AI 需要可验证的数据。海象在苏上解决了。
#287 🔗 原推
2026-09-22 00:00:08
Walrus
@SuiNetwork For agents to act autonomously in high-stakes environments, context and provenance aren't optional. Excited to show what we've been building at MBS next month.
中文: @SuiNetwork 为代理在高风险环境中自主行动，上下文和来源并非可选。很高兴能展示我们下个月在MBS所打造的内容。
#286 🔗 原推
2026-09-21 23:59:59
Walrus
RT @SmartCryptoNew1: 💡 @LighthouseWeb3 is partnering with @GoPlugin to push forward a verifiable memory layer for AI agents and applications, combining persistent, encrypted, and verifiable data across IPFS, @Filecoin, and @WalrusProtocol. 🔐 With #Plugin’s external data infrastructure, the collaboration explores how AI agents can securely store, access, verify, and act on data across decentralized networks—turning memory into a more reliable foundation for autonomous applications. ⚡ The goal is simple: give AI agents a memory layer they can trust, verify, and actually use. 🔽 VISIT https://plugin.global/ #SCN1
中文: RT @SmartCryptoNew1: @LighthouseWeb3 正与 @GoPlugin 合作，推动为人工智能代理和应用程序提供可验证的存储层，将IPFS、@Filecoin 和 @WalrusProtocol 之间的持久、加密和可验证数据相结合。 🔐 借助#Plugin的外部数据基础设施，该协作探索了人工智能代理如何安全地在去中心化网络中存储、访问、验证和处理数据，将内存转化为更可靠的自主应用基础。 ⚡ 目标很简单:为人工智能代理提供一个他们可以信任、验证并实际使用的存储层。 🔽 访问 #SCN1
media 0 共 1 项
#285 🔗 原推
2026-09-21 23:51:08
Walrus
RT @WalrusProtocol: Walrus Session 8 is LIVE: Chatbots That Remember → Add persistent state to your bot → Deploy it live → Share your before/after write-up Sep 18 – Oct 9 | $2,000 WAL prize pool. Details here: https://www.deepsurge.xyz/hackathons/c0141a4a-21be-4009-bc63-7c168608c849 https://twitter.com/WalrusProtocol/status/2101011483379585088/photo/1
中文: RT @Walrus 协议:海象会话8是实时的:记住的聊天机器人 → 为你的机器人添加持久状态 → 实时部署 → 分享你的撰写前/稿后 9月18日 – 10月9日 | 2000美元WAL奖金池。 详情请见:
media 0 共 1 项
#284 🔗 原推
2026-09-21 23:47:14
Walrus
ICYMI:@genspark_ai COO @sang_wen joined @chain_ofthought to talk persistent context: how their second brain handles memory decay, updating state on the fly, and capturing real-world room audio for agents. Proud to be a sponsor of @chain_ofthought this season. Episode link: https://youtu.be/ED3m6A7ScKA?si=qM-_kQe7okEcgmQr
中文: ICYMI:@genspark_ai 首席运营官 @sang_wen 与 @chain_Ofthought 合作，讨论持续存在的语境:他们的第二个大脑如何处理内存衰减、动态状态更新，以及为代理人员录制真实房间的音频。 为本赛季成为@chain_Ofthought的赞助商而感到自豪。 剧集链接:
#283 🔗 原推
2026-09-21 18:03:23
Walrus
RT @theonlywill: SUI BASECAMP SPEAKER ROSTER PART 2 💧 From @WalrusProtocol, @bluefinapp, @CurrentSUI, @WaterX_app and @tradeonhudi, here are more of the builders I’m most excited to hear from. These are the teams helping shape where Sui goes next and I’m excited to hear what they’re building. October 7–8 in Singapore. See you at Basecamp. ⤵️
中文: RT @theonlywill:SUI BASECAMP 发言人 ROSTER 第二部分 💧 来自@WalrusProtocol、@bluefinapp、@CurrentSUI、@WaterX_app 和 @tradeonhudi，以下是我最期待听到的更多构建器。 这些团队正在帮助塑造闣鱼下一步的走向，我很兴奋地听到他们正在打造的是什么。 10月7日至8日，新加坡。在Basecamp见你。⤵️
🎬
视频
#282 🔗 原推
2026-09-21 17:53:50
Walrus
RT @SuiNetwork: An agent is only as good as the data it can prove. 💎 @WalrusProtocol joins Sui Basecamp as a Diamond sponsor. Walrus is the data platform built for the demands of AI. It makes data portable, protected, and verifiable by anyone. Created by the ex-Meta engineers behind Sui. https://twitter.com/SuiNetwork/status/2101793031435526350/photo/1
中文: RT @SuiNetwork:一个代理仅能证明其数据的良好性。 @WalrusProtocol 加入 Sui Basecamp 担任钻石赞助商。 海象是专为人工智能需求而构建的数据平台。它使数据具有可移植性、可保护性并可验证性。由苏伊身后的前Meta工程师创建。
media 0 共 1 项
#281 🔗 原推
2026-09-21 15:45:11
Walrus
RT @EpochSui: @WalrusProtocol has enormous potential, yet very few people have explored what it can actually do. We want to help change that. Try https://names.epochsui.com/ and see how easy it is to build your own website, for anything, in just a few steps. We cover your Walrus storage costs for the first two years. All you have to do is let your imagination run. A project, a dapp, a personal site, a page for your business: Epoch Names puts the power of Walrus in your hands, and our MCP makes it simple. Build in minutes what used to take hours.
中文: RT @EpochSui:@WalrusProtocol 具有巨大潜力，但很少有人真正探索它能做些什么。 我们想帮助改变这一点。 尝试 一个项目、一个应用程序、一个个人网站，以及一个适合您企业的页面:大纪元名称将海象之力掌握在您手中，而我们的MCP使其变得简单。 几分钟内就建造过去需要数小时的时间。
#280 🔗 原推
2026-09-21 15:35:27
Walrus
You spend an entire D&D session slaying a dark wizard, and three sessions later, the AI Dungeon Master brings him back to life because it forgot he died. As one of our Walrus Session 7 winners, @MrRobotJi solved this by building Aetheris World Engine on Walrus Memory. Instead of relying on fuzzy text similarity, Aetheris writes game state to an immutable ledger on Walrus Mainnet: → NPC status (Deceased NPCs stay dead) → World-tick timelines (prevents timeline bugs) → SEAL-encrypted secret notes (keeps plot twists hidden from players) Read the article on @Medium: https://medium.com/@mrrobotji/the-dead-boss-who-wouldnt-stay-dead-how-we-built-an-immutable-tabletop-world-engine-on-walrus-a7651697df9c?postPublishedType=initial
中文: 你花了整整一次D&D治疗，杀死了一位黑巫师，三次之后，AI《地下城大师》让他重新安然无恃，因为那场赛会忘了他去世了。 作为我们的“海象七号”获奖者之一，@MrRobotJi 通过在 Walrus Memory 上构建 Aetheris World 引擎来解决这个问题。 与其依赖模糊的文本相似性，Aetheris 而是将游戏状态写入 Walrus Mainnet 上的不可变账本: → NPC 状态(已故的 NPC 保持不变) → 世界时间线(防止时间线漏洞) → SEAL加密秘密笔记(对玩家隐瞒情节转折) 请在@Medium上阅读文章:
media 0 共 1 项
#279 🔗 原推
2026-09-21 15:08:07
Walrus
RT @EdCriptoFi: Can @WalrusProtocol and Memory Walrus be your personal trainer? Building for Walrus Session #8 https://twitter.com/EdCriptoFi/status/2101679373027791192/photo/1
中文: RT @EdCriptoFi:@WalrusProtocol 和 Memory Walrus 能成为你的私人教练吗? 为海象会议而建 #8
media 0 共 1 项
#278 🔗 原推
2026-09-20 21:17:12
Walrus
RT @0x_why85: @WalrusProtocol That's a smart way to handle memory!
中文: RT @0x_why85:@WalrusProtocol 这是一种处理内存的明智方法!
#277 🔗 原推
2026-09-20 21:16:21
Walrus
RT @AA_RonOnChain: Already have a fun idea for Session 8 😄
中文: RT @AA_RonOnChain:已经为Session 8😄提供了一个有趣的创意
#276 🔗 原推
2026-09-18 18:18:42
Walrus
Walrus Session 8 is LIVE: Chatbots That Remember → Add persistent state to your bot → Deploy it live → Share your before/after write-up Sep 18 – Oct 9 | $2,000 WAL prize pool. Details here: https://www.deepsurge.xyz/hackathons/c0141a4a-21be-4009-bc63-7c168608c849 https://twitter.com/WalrusProtocol/status/2101011483379585088/photo/1
中文: 《海象8》直播:记住的聊天机器人 → 为你的机器人添加持久状态 → 实时部署 → 分享你的撰写前/稿后 9月18日 – 10月9日 | 2000美元WAL奖金池。 详情请见:
media 0 共 1 项
#275 🔗 原推
2026-09-18 18:12:27
Walrus
RT @Neuron_Edge: @WalrusProtocol This is a huge step for agent interoperability 🔥 Portable memory means agents can actually build continuity across tools and workflows instead of starting from zero every time.
#274 🔗 原推
2026-09-18 14:49:59
Walrus
RT @kakii120: The future of digital reputation is here. ⚡ @conso_xyz is building a Consumer Reputation Layer powered by @WalrusProtocol & @SuiNetwork. Your activity. Your reputation. Your next opportunity. 🐻‍❄️ https://twitter.com/kakii120/status/2100953192573862060/photo/1
media 0 共 1 项
#273 🔗 原推
2026-09-18 14:48:17
Walrus
RT @wal_notes: What happens if WalNotes disappears tomorrow? Your notes stay alive on @WalrusProtocol. We're open-sourcing an offline viewer so you can always fetch and decrypt your data straight from the network without us. 🌐 https://twitter.com/wal_notes/status/2100951525178654772/photo/1
中文: RT @wal_notes:如果明天 WalNotes 消失会发生什么? 你的笔记在@WalrusProtocol上保持鲜活。我们正在开源离线查看器，以便您无需我们，随时能直接从网络获取和解密您的数据。🌐
media 0 共 1 项
#272 🔗 原推
2026-09-18 14:48:16
Walrus
As one of our Walrus Session 7 winners, @neopsih upgraded Markov using Walrus Memory to solve how agent handoffs handle unverified state. When a fresh AI session blindly trusts notes from a prior run, it can easily resume work on a broken state. This upgrade turns handoffs into self-proving objects, forcing receiving agents to validate integrity and terminal receipts before accepting a state. Read on @Medium: https://medium.com/@anna.stolbovskaja/i-gave-a-fresh-agent-a-perfect-handoff-and-it-was-still-wrong-6020fa9501b5
中文: 作为我们的七号海象赛区获奖者之一，@neopsih 使用 Walrus Memory 升级了 Markov，以解决代理处理未经验证的状态问题。 当一个新的人工智能会话盲目地信任先前运行时的笔记时，它可以轻松地在状态崩溃时恢复工作。 此升级将接通方式转化为自证对象，迫使接收代理在接受状态前验证完整性和终端收据。 请在@Medium上阅读:
media 0 共 1 项
#271 🔗 原推
2026-09-18 11:29:00
Walrus
RT @ConorBronsdon: I'll be joining the Walrus Memory team for a private event with other founders & AI leaders in San Francisco next Thursday night to talk about the future of agentic commerce If you'd like an invite to join me + @WalrusProtocol & @SuiNetwork team next week - let me know! https://luma.com/sui-vf50
#270 🔗 原推
2026-09-18 04:33:23
Walrus
RT @theonlywill: SUI BASECAMP PROJECTS I’M MOST EXCITED TO HEAR FROM. 💧 @tradeonhudi, @WalrusProtocol, @bluefinapp, @AlphaFiSU, @CurrentSUI and @WaterX_app These are some of the builders helping shape the future of sui:native and I’m excited to hear news from them! October 7–8 in Singapore. See you at @SuiNetwork Basecamp. ⤵️
🎬
视频
#269 🔗 原推
2026-09-17 20:36:46
Walrus
RT @kimblgn: Quilt is one of the more underappreciated features of @WalrusProtocol. It can save builders a lot of engineering work and reduce their costs This is a great breakdown from BL showing how @_misofm is putting it to work
中文: RT @kimblgn:Quilt 是 @WalrusProtocol 中更被低估的功能之一。它可以为建筑工人节省大量工程工作并降低成本 这与BL的出色表现非常吻合，表明@_misofm正在将其付诸努力
#268 🔗 原推
2026-09-17 14:53:38
Walrus
Prompt history tells an agent what happened in a session. Memory gives it state that persists across runtimes, tools, and restarts. We're exploring persistent state execution at @HumanXCo on Sept. 23 at 3:00 p.m. CEST. Add @KostasKryptos Masterclass to your schedule. https://twitter.com/WalrusProtocol/status/2100599008540672210/photo/1
media 0 共 1 项
#267 🔗 原推
2026-09-17 14:53:25
Walrus
RT @EdCriptoFi: Memory Walrus Guide @walrusprotocol https://x.com/i/article/2100278441333473280
中文: RT @EdCriptoFi:记忆海象指南 @walrusprotocol
#266 🔗 原推
2026-09-16 22:19:36
Walrus
RT @lawrence_tilli: @kimblgn @PantsStanky @WalrusProtocol Couldn’t agree more. Glad to see it in such capable hands, and very excited to see what’s on the horizon. Show us the way @kimblgn! 🙌
中文: RT @lawrence_tilli:@kimblgn @PantsStanky @WalrusProtocol 无法再同意了。很高兴看到它如此能干，也非常兴奋地看到眼前的前景。 给我们看看 @kimblgn 的方式!🙌
#265 🔗 原推
2026-09-16 22:19:14
Walrus
RT @StatescuRazvan: Wrote up how Kraterion works, our Sui Overflow winner. It puts an S3-compatible API on top of Walrus Protocol and Sui. - Encryption by default (Seal threshold encryption) - Onchain access revocation - Persistent state via Walrus Memory Full breakdown on the blog: https://blog.walrus.xyz/kraterion-s3-storage-you-own-built-on-walrus-verifiable-agent-runtime/
#264 🔗 原推
2026-09-16 22:16:09
Walrus
Cloud storage locks agent memory in proprietary silos. If access drops, your agent's state vanishes. @StatescuRazvan built Kraterion to fix this: a drop-in S3 API using Walrus, Sui, and Seal threshold encryption for true data ownership. Read how he built it: https://blog.walrus.xyz/kraterion-s3-storage-you-own-built-on-walrus-verifiable-agent-runtime/
中文: 云存储将代理存储锁在专有的孤岛中。如果访问量下降，您的代理状态就会消失。 @StatescuRazvan 构建了 Kraterion 来修复此功能:使用 Walrus、Sui 和 Seal 的降落式 S3 API 进行真实数据所有权的阈值加密。 了解他是如何构建它的:
#263 🔗 原推
2026-09-16 22:16:06
Walrus
RT @kimblgn: For me, @WalrusProtocol is an untapped gold mine. I can see how relevant it could become to the problems emerging around AI We’ve only started turning those capabilities into products, and there’s so much more to build I think this is a huge opportunity
中文: RT @kimblgn:对我来说，@WalrusProtocol 是一个尚未开发的金矿。我能看出它与人工智能周围出现的问题是多么重要 我们才刚刚开始将这些能力转化为产品，而且还有更多需要建设的 我认为这是一个巨大的机遇
#262 🔗 原推
2026-09-16 18:36:55
Walrus
@Talus_Labs just released 5 specialized Coding Skills for agents, including native integration with Walrus so your workflows can handle state, task execution, and agentic memory right out of the box. Check out the repo and docs below:
中文: @Talus_Labs 刚刚发布了 5 个专为代理人员提供的专用编码技能，包括与 Walrus 的原生集成，让您的工作流程能够直接处理状态、任务执行和代理内存。 查看下面的文档和文档:
#261 🔗 原推
2026-09-16 15:18:05
Walrus
22 days until Sui Basecamp. What we’re bringing to Singapore will change how builders think about data & AI. Get ready for: - A whole new way to experience & manage agentic memory - Keynotes from @kostascrypto & @kimblgn - Live hands-on workshops, major partner showcases, and an unmissable reveal Catch us at the Diamond Booth Oct 7–8 at Marina Bay Sands. Big things are coming.
中文: 距离苏·巴斯坎普还有22天。 我们为新加坡带来的内容将改变建筑商对数据与人工智能的看法。准备好: - 一种全新的体验与管理特工记忆的方式 - 来自 @kostascrypto 和 @kimblgn 的主题 - 现场实践研讨会、主要合作伙伴展示以及不容错过的揭秘 10月7日至8日在滨海湾金沙码头的钻石展位上迎接我们。 大事即将到来。
#260 🔗 原推
2026-09-16 01:38:40
Walrus
RT @WalrusProtocol: @santiago_xvega @kostascrypto @HumanXCo Sí, la pérdida de contexto entre pasos destruye el flujo. Vamos a estar compartiendo el recap de la demo y todo lo de @HumanXCo la próxima semana aquí en X y en Linkedin para que estés al pendiente.
#259 🔗 原推
2026-09-15 20:59:29
Walrus
RT @0xmht: Thanks team 😊 This feels special coming from you. The CONSO × Walrus journey is just getting started, and we’re excited for what’s ahead. A lot more coming soon. Also excited to finally meet the team in person at Sui Basecamp in Singapore
中文: RT @0xmht:谢谢团队 😊 这从你身而来，感觉特别。 康索斯·沃勒斯之旅才刚刚开始，我们对未来的期待也非常期待。很快会有更多。 还很兴奋终于在新加坡苏大本营与团队见面
#258 🔗 原推
2026-09-15 20:56:32
Walrus
Congratulations to @conso_xyz who just crossed the 25,000+ users milestone! Excited to ship their upcoming products, incl. CONSO Mobile app & Consumer Passport and bring more users into the CONSO × Walrus ecosystem. Looking forward to achieving many more milestones together.
中文: 祝贺@conso_xyz，他刚刚突破了2.5万用户的里程碑! 对即将推出的产品感到兴奋，包括。CONSO 移动应用 & 消费者护照，并吸引更多用户进入 CONSO × Walrus 生态系统。 期待共同实现更多里程碑。
#257 🔗 原推
2026-09-15 17:45:00
Walrus
RT @wasimdropx: Really impressed with my experience using @conso_xyz and the Conso Extension. The ability to earn while using AI prompts, backed by @SuiNetwork & @WalrusProtocol, is a game changer for Web3 reputation layers.
中文: RT @wasimdropx:我使用 @conso_xyz 和 Conso 扩展程序的经验给我留下了深刻印象。使用由 @SuiNetwork & @WalrusProtocol 支持的人工智能提示来赚取收入，是 Web3 声誉层次的变革之作。
#256 🔗 原推
2026-09-15 15:07:13
Walrus
RT @sadboy_042: Thank you @WalrusProtocol for the spotlight Big thanks to @OlabanjiOlalek2 for the original prompt that sparked the idea 🙌 I built BuildMEM v2 to make context switching with Ai coding assistant a little less painful, with project-isolated memory, supersession tracking and preflight checks. Really cool seeing Walrus spotlight it. Appreciate you guys 🦭💙
中文: RT @sadboy_042:感谢@WalrusProtocol的关注 非常感谢@OlabanjiOlalek2 最初激发了这个想法的提示🙌 我构建了 BuildMEM v2，使使用 Ai 编程助手进行上下文切换时，只需将内存隔离、超会话跟踪和飞行前检查处理，缓解疼痛。 看到海象聚焦得非常酷。感谢你们🦭💙
#255 🔗 原推
2026-09-15 15:00:54
Walrus
RT @EpochSui: Supporting communities is how Epoch will feed the trenches. Onchain identity with .epoch names and a home on @WalrusProtocol , so every community on @SuiNetwork has a place to rally. The trenches built us. Now we build for them.
中文: RT @EpochSui:支持社区就是Epoch将如何滋养这些战壕。 带有 .epoch 名称的 Onchain 标识，以及 @WalrusProtocol 上的住宅，因此 @SuiNetwork 上的每个社区都有聚集的场所。 战壕造了我们。现在我们为他们而建。
#254 🔗 原推
2026-09-15 14:59:31
Walrus
RT @EmanAbio: TO THE #SUI COMMUNITY: basecamp is going to be a huge moment for all of us what we’re bringing to basecamp will change how people think about sui’s place in this industry this will go far beyond the technology. it will include things you’ve been asking for and things that matter directly to you this is going to be a turning point
中文: RT @EmanAbio:致#SUI社区:大本营将对我们所有人来说都是一个重要时刻 我们为大本营带来的内容将改变人们对我在这个行业中地位的看法 这将远远超出技术，包括你一直要求的东西以及对你而言重要的事情 这将是一个转折点
#253 🔗 原推
2026-09-15 14:37:35
Walrus
RT @kimblgn: If you’re heading to @HumanXCo, please make some time for this one @kostascrypto will be getting hands-on with Walrus Memory. It’s a good chance to try it, ask questions and explore what you could build
中文: RT @kimblgn:如果您要前往 @HumanXCo，请为此请花时间 @kostascrypto 将亲自动手关注 Walrus Memory。这是一个很好的机会，可以尝试一下，提出问题，并探索你能构建什么
#252 🔗 原推
2026-09-15 14:37:28
Walrus
Building a multiverse bedtime story generator for your kids is fun until your AI assumes that because "Pickles and Zorp" are best friends in Space World, they can't be strangers in Pirate World. Walrus Session 7 winner @AA_RonOnChain evolved Continuity Keeper on Walrus Memory to fix this parallel-canon bleed. His "Scoped Canon" model splits memory into CORE identity (who a character is) and REALITY layers (what happened in that world), letting characters explore alternate timelines without state collisions. Read the case study on @Medium: https://medium.com/@AA_RonOnChain/how-i-evolved-continuity-keeper-into-a-multiverse-storyteller-for-my-kids-fbab6be74938
中文: 为孩子打造一个多元的睡前故事生成器很有趣，直到你的人工智能认为“Pickles and Zorp”是《太空世界》中最好的朋友，因此在《海盗世界》中他们无法成为陌生人。 七号海象赛场冠军@AA_RonOnChain通过“持续守护者”来修复这种平行的混血。 他的“Scopeed Canon”模型将内存分割成核心身份(角色身份)和现实层次(即那个世界中发生的事情)，让角色在不发生状态碰撞的情况下探索不同的时间线。 请在@Medium上阅读案例研究:
media 0 共 1 项
#251 🔗 原推
2026-09-15 14:30:36
Walrus
@kostascrypto @HumanXCo local time in Amsterdam, Netherlands 🇳🇱
#250 🔗 原推
2026-09-15 14:09:01
Walrus
AI workflows move. Memory needs to move with them. @kostascrypto is taking that problem into a live Masterclass, demoing Walrus Memory at @HumanXCo on Sept. 23rd at 3pm (local time). Add it to your schedule. https://twitter.com/WalrusProtocol/status/2099824047659917427/photo/1
中文: 人工智能工作流程移动。记忆需要与它们一起移动。 @kostascrypto 正在将这一问题带入一个实时大师班，于9月23日下午3点(当地时间)在@HumanXCo上演示“海王记忆”。 将其添加到您的日程安排中。
media 0 共 1 项
#249 🔗 原推
2026-09-15 11:34:00
Walrus
Swapping sessions usually means your agent starts from zero or acts on outdated context. BuildMEM v2 fixes that. Built by Walrus Session 7 winner @sadboy_042 on Walrus Memory, it adds project-isolated namespaces, automatic supersession tracking, and preflight checks to Claude Code. Read the breakdown on @Medium: https://medium.com/@makindedaniel45/my-claude-code-sessions-kept-forgetting-everything-buildmem-v2-on-walrus-memory-14331ea069c2
media 0 共 1 项
#248 🔗 原推
2026-09-14 12:44:00
Walrus
Reputation systems run on continuous history. Great to see @conso_xyz using Walrus to keep consumer reputation data audit-ready and verifiable on Sui.
中文: 声誉系统基于连续的历史运行。 很高兴看到@conso_xyz 利用 Walrus 在 Sui 上保持消费者声誉数据审核的准备和可验证。
#247 🔗 原推
2026-09-12 17:50:53
Walrus
RT @conso_xyz: Just hit 25,000+ users on the CONSO Extension !! We’re delighted by the support from our community This milestone means a lot to us, and we’re just getting started More updates coming soon 👀
中文: RT @conso_xyz:刚在CONSO扩展上用户数量达到了2.5万以上! 我们对社区的支持感到非常高兴 这个里程碑对我们意义重大，而我们才刚刚开始 即将发布更多更新 👀
media 0 共 1 项
#246 🔗 原推
2026-09-12 17:50:34
Walrus
RT @Al_Nadim_nabil: @WalrusProtocol that's pretty cool how walrus memory keeps everything intact when switching models
中文: RT @Al_Nadim_nabil:@WalrusProtocol 真是太酷了，换个型号时，海象记忆如何保持了一切
#245 🔗 原推
2026-09-11 21:00:39
Walrus
No servers, no backend, no API keys, just your AI assistant drafting frontend code and serving the finished build directly on Walrus. Brilliant work by @EpochSui. 👏
中文: 无需服务器，没有后端，没有API密钥，只有您的AI助手在Walrus上直接编写前端代码并提供最终版本。 @EpochSui 的出色作品。👏
#244 🔗 原推
2026-09-11 20:20:00
Walrus
@SuiNetwork Already counting down to it.
中文: @SuiNetwork 已经对此进行了统计。
#243 🔗 原推
2026-09-11 20:04:28
Walrus
OpenAI’s new Agents API solves execution, but what about persistent context? Here is how to give your Agents API workflows portable, cross-session AND cross-platform context using Walrus Memory: https://walrus.xyz/products/walrus-memory/
中文: OpenAI 的新 Agents API 解决了执行问题，但持久性上下文又如何解决呢? 如何使用 Walrus Memory 为 Agents API 工作流程提供便携、跨平台和跨平台的环境:
#242 🔗 原推
2026-09-11 19:12:05
Walrus
RT @MPhuong609: @WalrusProtocol persistent agent memory sounds like a game changer, excited to see how it plays out at Basecamp
中文: RT @MPhuong609:@WalrusProtocol 持久的代理记忆听起来像是改变游戏规则，兴奋地看到它在 Basecamp 上的播放方式
#241 🔗 原推
2026-09-11 15:59:29
Walrus
RT @eightlends: @WalrusProtocol @HumanXCo @kostascrypto showing up with a real use case demo is how you actually make people care this should be obvious but ppl love theory way too much
中文: RT @eightlends:@WalrusProtocol @HumanXCo @kostascrypto 演示真实使用时，你真正让人得到关怀 这应该显而易见，但人们过于爱理论
#240 🔗 原推
2026-09-11 02:20:18
Walrus
If last year was great, this year will be even better. See you in Singapore. We're bringing persistent agent memory to Sui Basecamp.
中文: 如果去年是好的，今年会更好。 在新加坡见。我们为Sui Basecamp带来了持续的代理记忆。
#239 🔗 原推
2026-09-11 02:19:30
Walrus
If last year was the great, this year will be even better. See you in Singapore. We're bringing persistent agent memory to Sui Basecamp.
中文: 如果去年是伟大的，今年将会更好。 在新加坡见。我们为Sui Basecamp带来了持续的代理记忆。
#238 🔗 原推
2026-09-11 02:18:44
Walrus
If last year was the setup, this year is the upgrade. See you in Singapore. We're bringing persistent agent memory to Sui Basecamp.
#237 🔗 原推
2026-09-11 02:01:05
Walrus
RT @kimblgn: There is so much between an idea you really believe in and a product someone actually finds useful It’s easy to get excited about an idea like giving people more ownership of their data. I think the harder questions come when you have to decide what that should actually look like in a product What would someone use it for? How much would they need to understand to get something useful out of it? You start working through those details and realize how much the original idea left unanswered We’re working through a lot of this as we build products on Walrus There’s so much we want to make possible, and getting the products right is a huge part of that
#236 🔗 原推
2026-09-11 01:53:39
Walrus
RT @kimblgn: If AI becomes how people work, buy things and interact with businesses, a huge amount of economic activity will depend on these systems being able to access and use the right data Making that work across different businesses and AI systems is a big challenge in itself. That’s a big part of the opportunity I see for Walrus
#235 🔗 原推
2026-09-11 01:53:20
Walrus
RT @chaocacbannn: @WalrusProtocol @johnternus @Apple Persistent memory sounds promising, could really change the game for AI interactions
中文: RT @chaocacbannn:@WalrusProtocol @johnternus @Apple 持续内存听起来很有希望，可能彻底改变人工智能交互的模式
#234 🔗 原推
2026-09-10 21:20:04
Walrus
RT @EpochSui: Yesterday's @WalrusProtocol AMA turned into two promises for this week. Both shipped today. Expired hosting now tells the truth. The gateway used to answer "Site Not Found" and invite you to register a name that is already owned. Now it says what is actually true: the name is registered and owned, the content is not being served right now, and here is how to bring it back. A missing page inside a site that exists gets a real 404 instead of the same catch-all. You no longer find out from a dead site. Your Renewal tab now carries an amber counter that lights up about four months before hosting runs out, and it reads every blob of every site in one call instead of one scan per site. One thing we learned while building it, worth saying out loud: Walrus cannot extend storage that has already lapsed. "extend_blob" requires the blob to be certified and not expired, so a renewal bought after the fact would simply abort. We do not offer it. The tab tells you to publish the site again, and since Walrus blob ids come from the content, the same bytes give the same id and your name keeps pointing at it. We also fixed renewing from a second device, which quietly did not work. Your name never expires. The hosting does, and now you hear about it in time. https://names.epochsui.com/
中文: RT @EpochSui:昨天的@WalrusProtocol AMA本周变成了两个承诺。两者都今天发货了。 过期的托管现在说明了真相。用于回答“Site Nost Found”的网关，并邀请您注册已拥有的名称。现在它说明了事实:名称已注册并拥有，内容目前尚未送达，以下是将其重新发布的方式。存在的网站内缺失的页面会获得真正的404，而不是相同的全线。 你再也无法从死址中发现。您的续订选项卡现在会显示一个琥珀色计数器，该计数器在主机耗尽前大约四个月会亮起，它会将每个站点的每个点点读取一次，而不是每个站点的扫描。 我们在构建过程中了解到的一件事，值得大声说出来:海象无法延长已经失效的存储空间。“扩展”要求该漏洞必须经过认证且不会过期，因此事后购买的续期将干脆终止。我们不提供它。该标签页会要求您再次发布网站，由于Walrus blob的ID来自该内容，因此相同的字节会给出相同的ID，且您的名字会持续指向它。 我们还修复了第二台设备的续订，但该设备悄然无法使用。 你的名字永远不会过期。主机确实如此，现在你及时听到了。
#233 🔗 原推
2026-09-10 20:44:44
Walrus
RT @kostascrypto: Your Memory
中文: RT @kostascrypto:你的内存
#232 🔗 原推
2026-09-10 20:31:00
Walrus
RT @CarrySui: Four symptoms of the same problem: AI agents can’t carry trusted state across session and workflows Losing context is the cause. Repeating work and broken workflows are the symptoms. But having to start over is where users really feel it. Carry gives persistent, permissioned memory for AI agents, with verifiable receipts for every memory-based answers including refusals. @WalrusProtocol
中文: RT @CarrySui:同一问题的四个症状:人工智能代理无法跨会话和工作流程承载可信状态 失去背景是原因。重复工作和工作流程不断是症状。但必须重新开始，是用户真正感受到它的地方。 为AI代理提供持久且可许可的内存，并为包含拒绝在内的每个基于内存的回答提供可验证的收据。 @华勒斯协议
media 0 共 1 项
#231 🔗 原推
2026-09-10 20:18:06
Walrus
RT @WalrusProtocol: When an AI agent loses state, where do you feel it first? Let us know in the poll below. We're taking the question into @HumanXco. 🇳🇱
中文: RT @WalrusProtocol:当人工智能代理失去状态时，你首先会从哪里感受到? 请在下方的投票中告诉我们。 我们正在将问题纳入@HumanXco。🇳🇱
#230 🔗 原推
2026-09-10 14:36:47
Walrus
"Trust only goes so far when your data is no longer yours to control." — @johnternus @Apple's right: cloud memory needs blind trust. On-device means agents reset every session. Enter Walrus Memory: persistent, user-controlled memory for AI agents. https://www.youtube.com/live/39BalPDuTo0?si=MeCXaa2XmmAGoitx&t=320
中文: 信任只有在数据不再由你掌控时才会持续。——@johnternus @Apple 的没错:云内存需要盲目信任。设备上意味着代理会重置每个会话。 输入海象记忆:用于人工智能代理的持久、用户控制的内存。
#229 🔗 原推
2026-09-09 23:21:26
Walrus
"Trust only goes so far when your data is no longer yours to control." — @johnternus, Apple @Apple's right: handing your context to centralized cloud servers requires blind trust. But keeping it locked on-device means your #AI agents reset the moment a session ends. That's why we built Walrus Memory: persistent, user-controlled memory for AI agents. https://www.youtube.com/live/39BalPDuTo0?si=MeCXaa2XmmAGoitx&t=320
中文: 信任只有在你的数据不再由你掌控时才会持续。——@johnternus，苹果 @Apple 的右:将您的上下文交给集中式云服务器需要盲目信任。 但将设备锁定在设备上，意味着你的#AI代理会重置会话结束的瞬间。 这就是我们为人工智能代理构建“海象记忆”的原因:即持久且由用户控制的内存。
#228 🔗 原推
2026-09-09 23:14:58
Walrus
RT @kimblgn: I can see more people asking what happens to the data they share with AI and how much control they actually have over it. I think those questions are only going to get bigger as AI becomes more and more useful We’re going to give these systems a lot of context about our lives and businesses. Over time, that could become some of the most useful knowledge we have. And I want people to have a real say in how it’s used We’ve already made some of this possible with Walrus Memory. And I think there’s a lot more we can build around the same idea using @WalrusProtocol as a foundation
中文: RT @kimblgn:我可以看到更多人在询问他们与人工智能共享的数据会发生什么，以及他们实际对这些数据拥有多少控制。我认为随着人工智能变得越来越有用，这些问题只会变得越来越大 我们将为这些系统提供大量关于我们生活和业务的背景。随着时间的推移，这可能成为我们掌握的最有用的知识之一。我希望人们能真正地对它的使用方式拥有发言权 我们已经通过《海象记忆》实现了一些这样的可能。我认为，使用@WalrusProtocol作为基础，我们可以围绕同一个理念进行更多的构建
#227 🔗 原推
2026-09-09 21:40:50
Walrus
RT @EpochSui: What an AMA! Thank you @WalrusProtocol for the warm welcome, felt like home! Shoutout to @BiasGoose for hosting and to everyone who brought amazing questions, hope I delivered! I'm always available for any other questions, feel free to reach out anytime! And one more thing… big Epoch Pay updates dropping in the coming days
中文: RT @EpochSui:真是个美国医学会!谢谢@WalrusProtocol，热情欢迎，感觉宾至如一日! 向@BiasGoose大声呼喊，感谢他们的主持，以及所有带来精彩疑问的人，希望我能送上场! 我随时可以回答其他问题，随时可以联系! 还有一件事......未来几天大的大纪元支付动态将下降
#226 🔗 原推
2026-09-09 21:39:55
Walrus
RT @cryptosetters: @WalrusProtocol @Sceat_ Looking forward to seeing more builders from the Walrus community! 🔥
中文: RT @cryptosetters:@WalrusProtocol @Sceat_ 期待看到更多来自 Walrus 社区的构建者!🔥
#225 🔗 原推
2026-09-09 21:39:49
Walrus
RT @cryptosetters: @WalrusProtocol @discord @EpochSui My favorite Walrus 🦭💙 Always building, always innovating. Walrus season! 🔥
中文: RT @cryptosetters:@WalrusProtocol @discord @EpochSui 我最喜欢的海象 🦭💙 始终不断建设，不断创新。海象季节!🔥
#224 🔗 原推
2026-09-09 21:39:32
Walrus
RT @EdCriptoFi: @EpochSui @WalrusProtocol @BiasGoose Great AMA, thx for answering my questions
中文: RT @EdCriptoFi:@EpochSui @WalrusProtocol @BiasGoose 很棒 AMA，供我解答
#223 🔗 原推
2026-09-09 21:39:26
Walrus
RT @Sceat_: Try it with your agent, @suize_io ! It's as fast as vercel
中文: RT @Sceat_:请与您的代理 @suize_io 一起试用!速度和 vercel 一样快
#222 🔗 原推
2026-09-09 14:27:40
Walrus
RT @suize_io: Fast websites, public or private, and even with your custom domain! See how @Sceat_ unlocked walrus sites for the agentic web
中文: RT @suize_io:快速网站，无论是公开还是私密，甚至使用您的自定义域名!了解 @Sceat_ 如何为代理网络解锁 walrus 网站
#221 🔗 原推
2026-09-09 14:27:25
Walrus
RT @delf002: @iamkjess @SuiNetwork @conso_xyz @WalrusProtocol @EpochSui See in you discord for the Epoch AMA
中文: RT @delf002:@iamkjess @SuiNetwork @conso_xyz @WalrusProtocol @EpochSui 在你这里为大纪元AMA寻找不和
#220 🔗 原推
2026-09-09 14:26:10
Walrus
We're 45 minutes away from our @Discord AMA with @EpochSui. Bring your questions, we're diving into each one. Discord: https://discord.com/invite/mBKgWWcB4p
中文: 我们距离@Discord AMA 有 45 分钟车程，@EpochSui。 带上你的问题，我们深入探讨每一个问题。 不和:
#219 🔗 原推
2026-09-09 14:22:48
Walrus
RT @WalrusProtocol: When an AI agent loses state, where do you feel it first? Let us know in the poll below. We're taking the question into @HumanXco. 🇳🇱
中文: RT @WalrusProtocol:当人工智能代理失去状态时，你首先会从哪里感受到? 请在下方的投票中告诉我们。 我们正在将问题纳入@HumanXco。🇳🇱
#218 🔗 原推
2026-09-09 14:19:01
Walrus
We're 45 minutes away from our @Discord AMA with @EpochSui. Bring your questions, we're diving into each one. Walrus Discord: https://discord.com/invite/mBKgWWcB4p
中文: 我们距离@Discord AMA 有 45 分钟车程，@EpochSui。 带上你的问题，我们深入探讨每一个问题。 海象对冲:
#217 🔗 原推
2026-09-09 14:15:39
Walrus
RT @SuiCommunity_ID: 🚨 Builder AMA bersama @EpochSui ! Besok, 9 September pukul 22:00 WIB, Walrus akan mengadakan AMA bersama @EpochSui! 📍 Join di Discord: https://discord.com/channels/946098997637042178/1486488125696905438 💬 Punya pertanyaan untuk Epoch? Kirim pertanyaan terbaikmu dan berkesempatan mendapatkan merchandise! See you 🫰 https://twitter.com/SuiCommunity_ID/status/2097429903402795063/photo/1
media 0 共 1 项
#216 🔗 原推
2026-09-09 03:32:56
Walrus
RT @EpochSui: Tomorrow, 3 PM UTC, Walrus Discord. Bring your questions, we'll be there!
中文: RT @EpochSui:明天，UTC下午3点，海象对应。带上你的问题，我们会到的!
#215 🔗 原推
2026-09-09 03:32:52
Walrus
RT @EdCriptoFi: Alpha
#214 🔗 原推
2026-09-09 03:32:45
Walrus
RT @WalrusProtocol: Join us tomorrow Sept 9th at 3:00 PM UTC on Walrus' @discord for an AMA with Stefano, founder of @EpochSui. We'll dive into trust infrastructure on @SuiNetwork, trustless vesting, and .epoch names. Join here: https://discord.com/invite/mBKgWWcB4p https://twitter.com/WalrusProtocol/status/2097439584787611792/photo/1
media 0 共 1 项
#213 🔗 原推
2026-09-09 03:32:39
Walrus
RT @cryptosetters: @WalrusProtocol @Sceat_ My all-time favorite kind of builder spotlight. AI agents that can go from code to deployment without the usual friction — this is the future. Walrus is making it happen. 🔥
中文: RT @cryptosetters:@WalrusProtocol @Sceat_ 我最喜爱的建筑工人聚光灯。能够在不产生常规摩擦的情况下从代码转到部署的人工智能代理——这就是未来。海象正在让它成为现实。🔥
#212 🔗 原推
2026-09-09 03:30:06
Walrus
Join us tomorrow Sept 9th at 3:00 PM UTC on Walrus' @discord for an AMA with Stefano, founder of @EpochSui. We'll dive into trust infrastructure on @SuiNetwork, trustless vesting, and .epoch names. Join here: https://discord.com/invite/mBKgWWcB4p https://twitter.com/WalrusProtocol/status/2097439584787611792/photo/1
中文: 明天9月9日，即美国时东时节下午3:00，与@EpochSui创始人斯特凡诺共同参加华磕的@discord。 我们将深入探讨@SuiNetwork、无信任背带和.epoch名称的信任基础设施。 加入此处:
media 0 共 1 项
#211 🔗 原推
2026-09-08 21:39:00
Walrus
Read the full Build Note to see how @Sceat_ won the Sui Overflow hackathon and built an agent-native publishing flow: https://blog.walrus.xyz/suize-agent-ships-walrus-website/ 🏅
中文: 阅读完整的 Build Note 内容，了解 @Sceat_ 如何赢得 Sui Overflow 黑客马拉松，并构建了代理原生出版流程: 🏅
#210 🔗 原推
2026-09-08 19:24:53
Walrus
How @Sceat_ engineered it: Bundles site assets into a single package to keep storage efficient. Uses fast edge caching so sites load instantly. Guarantees the site stays online independently without relying on a central company server.
中文: @Sceat_ 是如何设计的: 将站点资产捆绑成单一套餐，以确保存储效率。 使用快速边缘缓存，以便网站立即加载。 保证网站在不依赖中央公司服务器的情况下保持在线状态。
#209 🔗 原推
2026-09-08 19:24:53
Walrus
🧵 1/4: Builder Spotlight: @Sceat_ 🛠 You give Claude Code or Cursor a prompt, it writes the code, fixes its errors, and drops a completed project folder in front of you. Then it stops; because publishing a website usually requires human accounts, credit cards, and logins. @Sceat_ built Suize to fix the last step:
media 0 共 1 项
#208 🔗 原推
2026-09-08 19:24:52
Walrus
When an AI agent loses state, where do you feel it first? Let us know in the poll below. We're taking the question into @HumanXco. 🇳🇱
中文: 当人工智能代理失去状态时，你首先会从哪里感受到? 请在下方的投票中告诉我们。 我们正在将问题纳入@HumanXco。🇳🇱
#207 🔗 原推
2026-09-08 15:50:56
Walrus
Heading to @HumanXCo in Amsterdam? 🇳🇱 We’re joining 8,000+ AI builders to solve a critical production issue: agents that reset every time a session ends. Catch @kostascrypto on Day 1 for our interactive Masterclass:"Memory that survives the session: building agents that don't start from zero" No slides, just a live demo on wiring portable memory into your agents and solving real failure scenarios live. Make sure to add it to your schedule.
#206 🔗 原推
2026-09-08 15:48:10
Walrus
RT @HumanXCo: The HumanX app is now live. Attendees are messaging, browsing the community, and booking 1:1 meetings— the earlier you're in there, the better your week in Amsterdam looks. Not registered? Save €500 by midnight on Friday, 4 September 👉 https://hubs.la/Q04wf2-l0 #HumanX #AI https://twitter.com/HumanXCo/status/2095092603012182257/video/1
中文: RT @HumanXCo:HumanX 应用程序现已上线。参会者是即时通讯、浏览社区以及预订1:1会议——越早进入阿姆斯特丹，你的一周观看时间就越好。 未注册?9月4日星期五午夜前节省500欧元 👉 #HumanX #AI
🎬
视频
#205 🔗 原推
2026-09-08 15:45:43
Walrus
RT @HumanXCo: The countdown is on! Less than one month until HumanX Amsterdam 2026 🎉The EU AI Act’s transparency rules are now in force. This is where the strategy for staying compliant and competitive gets worked out. €500 saving ends Friday 4 September https://hubs.la/Q04v9g680 #HumanX https://twitter.com/HumanXCo/status/2092191295439561141/video/1
中文: RT @HumanXCo:倒计时已开始!距离《欧洲国际公共新闻》《阿姆斯特丹2026年法案》的透明度规定生效不到一个月。这就是保持合规和竞争策略的制定。 500欧元储蓄，9月4日星期五结束 #HumanX
🎬
视频
#204 🔗 原推
2026-09-08 15:38:35
Walrus
RT @kimblgn: This is where the world is heading. AI companies are racing to build the system that knows you the best To do that, they need access to the data you generate through your life and work. Which is all private & sensitive. That data gives it the context to understand what matters to you and why My view is that it should belong to you. You should decide which models and applications can access it, and carry it with you as those tools change This is the broader vision guiding our work at @WalrusProtocol
中文: RT @kimblgn:这就是世界的发展方向。人工智能公司正在争相打造最了解你的系统 要做到这一点，他们需要获取你通过生活和工作生成的数据。哪个都是私密的和敏感的。这些数据为它提供了理解你重要哪些内容以及原因的背景 我认为它应该属于你。您应该决定哪些模型和应用程序可以访问它，并在这些工具发生变化时随身携带 这是指导我们工作的更广泛愿景，支持@WalrusProtocol
#203 🔗 原推
2026-09-07 21:46:58
Walrus
A hotel search yielding stale rates is bad state management. Exposing another user's prompt is a lack of access control. If memory isn't encrypted, isolated, and verifiably updated at the storage layer, agents will keep leaking private context and serving stale outputs.
#202 🔗 原推
2026-09-07 21:45:38
Walrus
Next stop: Sui Basecamp in Singapore. 🇸🇬 We'll be talking about state persistence, portable memory, and where agent stacks are going. Make your card and drop it below so we can see who's on the ground. https://cards.buildonsui.io/c/walrusprotocol
中文: 下一站:新加坡的苏大本营。🇸🇬 我们将讨论状态持久化、可移植内存以及代理堆栈的去向。 制作你的卡片，然后放在下方，以便我们看看谁在地面上。
#201 🔗 原推
2026-09-06 17:14:12
Walrus
Next stop: Sui Basecamp in Singapore. 🇸🇬 We'll be talking about state persistence, portable memory, and where agent stacks are going. Make your card and drop it below so we can see who's on the ground. https://cards.buildonsui.io/
中文: 下一站:新加坡的苏大本营。🇸🇬 我们将讨论状态持久化、可移植内存以及代理堆栈的去向。 制作你的卡片，然后放在下方，以便我们看看谁在地面上。
#200 🔗 原推
2026-09-06 17:12:08
Walrus
RT @Ucaird_zenith: @WalrusProtocol Portable memory? Love that, Walrus!
中文: RT @Ucaird_zenith:@WalrusProtocol 便携式内存?太喜欢了，海象!
#199 🔗 原推
2026-09-06 16:15:09
Walrus
RT @WalrusProtocol: Where should you actually store AI agent data? No single database model solves every problem. Vectors are not your source of truth, secrets are not memory, and raw chat logs do not scale. In our latest post, @JessieWritesx breaks down how to match each class of agent data to the store built for it. Read the full breakdown: https://blog.walrus.xyz/where-to-store-ai-agent-data/
中文: RT @WalrusProtocol:你究竟应该在哪里存储人工智能代理数据? 没有单一的数据库模型能解决每一个问题。矢量不是你的真理来源，秘密不是记忆，而原始的聊天记录不会扩展。 在我们最新的帖子中，@JessieWritesx 详细列出了如何将每类代理数据与为其构建的存储方式匹配。 阅读完整详情:
media 0 共 1 项
#198 🔗 原推
2026-09-06 16:13:32
Walrus
RT @ariell_xyz: @WalrusProtocol portable memory under your control sounds exactly right
中文: RT @ariell_xyz:@WalrusProtocol 可移植内存在您掌控下，音效完全正确
#197 🔗 原推
2026-09-05 17:24:20
Walrus
RT @ariell_xyz: @WalrusProtocol portable memory across sessions sounds really useful
中文: RT @ariell_xyz:@WalrusProtocol 跨会话的便携记忆听起来非常有用
#196 🔗 原推
2026-09-05 17:21:44
Walrus
RT @Ucaird_zenith: @WalrusProtocol Nice that Walrus Memory keeps context =))
中文: RT @Ucaird_zenith:@WalrusProtocol Nice 认为 Walrus 记忆保持了上下文 =)
#195 🔗 原推
2026-09-05 17:21:34
Walrus
RT @NBCryptodotsui: October is loading… 👀 From @SuiNetwork to @ikadotxyz , @WalrusProtocol , and the wave of new DeFi protocols entering alpha it feels like everything is lining up. Only a few days left. This month, Sui ecosystem is about to hit different. 🌊 Watch the wave. 🫡💧🦕🦑
media 0 共 1 项
#194 🔗 原推
2026-09-04 15:04:49
Walrus
RT @lakshay: been using @conso_xyz today n its actually a pretty neat idea lol. every day we're asking AI to search, code, write, learn n solve problems... those interactions create value, but users usually dont get anything back. CONSO flips that by rewarding your everyday AI usage through its browser extension. connected my ChatGPT, Claude, Gemini n Perplexity in like 2 mins n already started stacking ZAPs ⚡️ feels nice earning a little something for prompts i was gonna make anyway .... if u wanna try it: 🔗 https://conso.xyz/extension ref code: CONSO-H8OGU @conso_xyz https://x.com/conso_xyz/status/2094651000703660201/video/1
media 0 共 2 项
🎬
视频
#193 🔗 原推
2026-09-04 14:58:47
Walrus
RT @wal_notes: What happens if WalNotes disappears tomorrow? Your notes stay alive on @WalrusProtocol. We're open-sourcing an offline viewer so you can always fetch and decrypt your data straight from the network without us. 🌐 https://twitter.com/wal_notes/status/2095855663473709523/photo/1
中文: RT @wal_notes:如果明天 WalNotes 消失会发生什么? 你的笔记在@WalrusProtocol上保持鲜活。我们正在开源离线查看器，以便您无需我们，随时能直接从网络获取和解密您的数据。🌐
media 0 共 1 项
#192 🔗 原推
2026-09-04 14:46:53
Walrus
RT @cryptosetters: @WalrusProtocol This is really useful for AI agents. Memory that can move across apps and tools is a big step forward. Walrus is building something special.
中文: RT @cryptosetters:@WalrusProtocol 这对人工智能代理非常有用。 能够跨应用和工具移动的内存是向前迈出的一大步。海象正在建造一些特别的东西。
#191 🔗 原推
2026-09-04 14:46:31
Walrus
Big updates coming to Singapore. 🇸🇬 Join our Head of Product @kimblgn and the rest of the Walrus team at Sui Basecamp on Oct 7-8. Passes still available: https://sui.io/basecamp
中文: 重大消息即将进入新加坡。🇸🇬 10月7日至8日，请加入我们的产品主管@kimblgn以及海象队的其他成员。 仍有通行证可供使用:
#190 🔗 原推
2026-09-04 14:37:38
Walrus
RT @kimblgn: The AI revolution is also a data revolution Few understand this
中文: RT @kimblgn:人工智能革命也是一场数据革命 很少有人理解这一点
#189 🔗 原推
2026-09-04 14:17:23
Walrus
RT @kimblgn: I have a lot to share at @SuiNetwork Basecamp The bigger vision for @WalrusProtocol. What Walrus Memory has already proven. And something new our team has been building Can’t wait https://twitter.com/kimblgn/status/2095875780958154906/photo/1
中文: RT @kimblgn:我在@SuiNetwork Basecamp 有很多内容要分享 @WalrusProtocol 的更大愿景。海象记忆已经证明了这一点。而我们团队一直在构建一些新东西 迫不及待
media 0 共 1 项
#188 🔗 原推
2026-09-04 14:17:19
Walrus
RT @DavidTiczon: Who else is coming to @SuiNetwork 💧 Basecamp this year? https://twitter.com/DavidTiczon/status/2095551654024348043/photo/1
中文: RT @DavidTiczon:今年还有谁来参加@SuiNetwork 💧 Basecamp?
media 0 共 1 项
#187 🔗 原推
2026-09-03 20:58:30
Walrus
RT @EvanWeb3: I'm heading to Sui Basecamp 2026. Make your own card and let everyone know you'll be there 👇 https://cards.buildonsui.io/c/evanweb3
中文: RT @EvanWeb3:我将前往苏大本营2026。自己制作卡片，让大家知道你会到那里 👇
#186 🔗 原推
2026-09-03 20:58:15
Walrus
RT @SuiGkim: Well, this just made Basecamp even bigger. @EvanWeb3 will be in there! 🔥
中文: RT @SuiGkim:这让Basecamp变得更大了。 @EvanWeb3 会在那里!🔥
#185 🔗 原推
2026-09-03 20:57:39
Walrus
RT @0xSHUBY: @WalrusProtocol @conso_xyz Bullish on walrus from the start and on early access of conso
中文: RT @0xSHUBY:@WalrusProtocol @conso_xyz 从一开始就在海象上，也在早期访问Conso上
#184 🔗 原推
2026-09-03 20:56:01
Walrus
RT @sheikhakash69: @WalrusProtocol Portable memory across tools is exactly the missing piece for agents.
中文: RT @sheikhakash69:@WalrusProtocol 跨工具的便携式内存恰好是代理人员缺失的部分。
#183 🔗 原推
2026-09-03 20:37:18
Walrus
RT @wal_notes: If we can't read your private notes, who can? Only you. With client-side encryption via your Sui wallet, not even our team or Walrus node operators can peek at your thoughts. Your keys, your brain. 🔐🦭
中文: RT @wal_notes:如果我们无法阅读您的私人笔记，谁可以?只有你。 通过您的Sui钱包进行客户端加密，甚至连我们的团队或Walrus节点运营商都无法窥视您的想法。钥匙，你的大脑。🔐🦭
#182 🔗 原推
2026-09-03 14:39:33
Walrus
Where should you actually store AI agent data? No single database model solves every problem. Vectors are not your source of truth, secrets are not memory, and raw chat logs do not scale. In our latest post, @JessieWritesx breaks down how to match each class of agent data to the store built for it. Read the full breakdown: https://blog.walrus.xyz/where-to-store-ai-agent-data/
中文: 你究竟应该在哪里存储人工智能代理数据? 没有单一的数据库模型能解决每一个问题。矢量不是你的真理来源，秘密不是记忆，而原始的聊天记录不会扩展。 在我们最新的帖子中，@JessieWritesx 详细列出了如何将每类代理数据与为其构建的存储方式匹配。 阅读完整详情:
media 0 共 1 项
#181 🔗 原推
2026-09-03 14:28:52
Walrus
RT @EdCriptoFi: The chaos is organized. Walnotes is live. Explore it now. 🗂️ @wal_notes Create your free account now: https://walnotes.xyz/
中文: RT @EdCriptoFi:混乱已经形成。 华尔诺特是活的。 立即探索。🗂️ @wal_notes 立即创建您的免费账户:
#180 🔗 原推
2026-09-03 00:10:57
Walrus
RT @MeniyaAnilYT: Best Way To Get The Max Zaps On @conso_xyz ? Just Do Your Genuine Everyday Research On Claude & ChatGPT, Keep All Platforms Connected, And Maintain Your Daily Streak. @WalrusProtocol Earning Rewards While Actually Getting Work Done Hits Different 🐻 https://twitter.com/MeniyaAnilYT/status/2095175080989044843/video/1
🎬
视频
#179 🔗 原推
2026-09-03 00:08:49
Walrus
RT @MeniyaAnilYT: How @conso_xyz Actually Works In 3 Simple Steps: 1. Send A Prompt To AI (ChatGPT, Claude, Gemini, Perplexity) 2. Earn Zaps Automatically For Your Activity 3. Verified & Powered By @WalrusProtocol More Activity = More Zaps. Your Daily AI Workflow Now Has Real Value 🐻⚡ #Web3‌‌ https://twitter.com/MeniyaAnilYT/status/2095182187280879806/photo/1
中文: RT @MeniyaAnilYT:@conso_xyz 如何通过三个简单步骤实际工作: 1。向人工智能发送提示(ChatGPT、克劳德、双子座、困惑) 2。为您的活动自动赚取 Zaps 3。已验证和支持;由 @WalrusProtocol 提供支持 活动量 = 更多 Zaps。您的每日AI工作流现在具有实际价值 🐻⚡ #Web3
media 0 共 1 项
#178 🔗 原推
2026-09-03 00:08:17
Walrus
"Who is the primary user beyond the Walrus ecosystem, and what problem does WalNotes solve better than established offline-first tools?" Great question coming out of today’s AMA with @Edcriptofi, by @BludAndrea. Secure storage shouldn't feel like homework. @wal_notes brings zero-friction note-taking powered by Walrus, giving users complete data ownership and verifiable privacy without the setup headaches. 🦭
中文: 除了海象生态系统之外，谁是主要用户?WalNotes 比已建立的离线优先工具更能解决什么问题? 今天与@Edcriptofi合作的AMA，由@BludAndrea提出，这是个绝佳的问题。 安全存储不应像作业一样。@wal_notes 采用 Walrus 支持的零摩擦笔记，为用户提供完整的数据所有权和可验证的隐私保护，且无需设置问题。🦭
#177 🔗 原推
2026-09-03 00:01:48
Walrus
Thanks to everyone who joined our AMA Walrus @discord today for a conversation led by @EdCriptoFi on @wal_notes. Great to see everyone engaged, asking questions, in community.
中文: 感谢今天加入我们的AMA Walrus @discord的所有人，他们通过@EdCriptoFi在@wal_notes上进行了对话。 很高兴看到大家积极参与，在社区里提出问题。
#176 🔗 原推
2026-09-02 23:55:16
Walrus
RT @Just1Sui: The whole justonesui site now runs on Walrus at onesui.epoch. Home, the lore, 56 community memes and the live burn feed. Every image on Walrus too. No server behind any of it. #EpochSui @EpochSui https://onesui.epochsui.com/ https://twitter.com/Just1Sui/status/2095245532432105810/photo/1
中文: RT @Just1Sui:整个 justonesui 网站现在在 onesui.epoch 的 Walrus 上运行。 家、传说、56个社区表情包和活体燃烧饲料。瓦鲁斯上的每一张图片。没有任何服务器。 #EpochSui @EpochSui
media 0 共 1 项
#175 🔗 原推
2026-09-02 23:51:24
Walrus
RT @EdCriptoFi: GM ☕️ , Big day ahead
中文: RT @EdCriptoFi:通用汽车☕️，未来重要一天
#174 🔗 原推
2026-09-02 14:59:29
Walrus
RT @wal_notes: GM, The chaos is organized. @wal_notes is live. Explore it now. 🗂️ https://walnotes.xyz/ https://twitter.com/wal_notes/status/2095114817564533081/photo/1
中文: RT @wal_notes:总经理，混乱是有组织的。 @wal_notes 是实时的。 立即探索。🗂️
media 0 共 1 项
#173 🔗 原推
2026-09-02 14:59:27
Walrus
RT @grossbel12: @WalrusProtocol @EdCriptoFi @wal_notes @discord I am going!!! Do not miss guys!
中文: RT @grossbel12:@WalrusProtocol @EdCriptoFi @wal_notes @discord 我要去!!!不要错过!
#172 🔗 原推
2026-09-02 14:59:23
Walrus
RT @lyraai_space: Every action leaves a receipt on the chain. With Walrus, those receipts stick around and can't be altered. You can always check what it did. No more guessing. @WalrusProtocol @SuiNetwork
中文: RT @lyraai_space:每个操作都会在链条上留下一张收据。 使用海象，这些收据会一直存在，无法更改。 你随时都可以查看它做了什么。不再猜测。 @WalrusProtocol @SuiNetwork
#171 🔗 原推
2026-09-02 14:59:17
Walrus
RT @conso_xyz: @WalrusProtocol Great to be building alongside Walrus 🐻‍❄️🤝 Prompt to Earn season has begun.
中文: RT @conso_xyz:@WalrusProtocol 与 Walrus 并列 🐻 ❄️🤝 快速赚钱的季节已经开始了。
#170 🔗 原推
2026-09-02 14:58:48
Walrus
RT @conso_xyz: The CONSO Extension turns your everyday AI activity into Stats, Insights and Rewards. With it, you can: – Track prompts, tokens, models and estimated API cost across the Top AI platforms – Understand your prompting patterns and quality – Unlock achievements and climb leaderboard – Participate in smart campaigns & missions – Discover AI products looking for real users We never collect or store your conversations. The future of AI shouldn’t just be about models getting better, You should participate in the value you create. CONSO helps you track it, and earn your share 🐻‍❄️⚡️
中文: RT @conso_xyz:CONSO扩展将你的日常人工智能活动转化为统计、洞察和奖励。 有了它，你可以: – 跟踪顶级人工智能平台的提示、令牌、模型和估计API成本 了解你的提示模式和质量 - 解锁成就并攀爬排行榜 – 参与智能活动与任务 – 发现寻找真实用户的人工智能产品 我们从不收集或存储您的对话。 人工智能的未来不应仅仅与模型的改善有关，而应参与你创造的价值。 CONSO 可帮助您追踪并赚取份额 🐻 ❄️⚡️
🎬
视频
#169 🔗 原推
2026-09-02 14:55:01
Walrus
RT @conso_xyz: Introducing the CONSO Extension - Prompt to Earn, turning your everyday AI activity into real rewards 🐻‍❄️⚡️ Every day, millions of people use AI to search, create, code, learn and make decisions But there’s a hidden trade: - You pay for the models - Your interactions make those models more valuable - Yet you get almost nothing back from the value you create CONSO flips that equation https://conso.xyz/extension
🎬
视频
#168 🔗 原推
2026-09-02 14:53:32
Walrus
AI activity is becoming a new layer of digital reputation. @conso_xyz is turning that activity into a track record users can own and benefit from, powered by Walrus. Prompt to Earn is now LIVE. 🐻‍❄️⚡️
#167 🔗 原推
2026-09-02 14:51:58
Walrus
RT @thedevranjan: ✨I'm #10963 in line for @conso_xyz's Zap Reactor Early Access ⚡️ Powered by @SuiNetwork & @WalrusProtocol Only 15,000 spots, Go claim yours : https://conso.xyz/access Conso is your AI use private tracker which does not leave your out of your computer and stays within your browser. Real insights into your prompts, platforms, sessions, and habits. Analyze your activity and improve the way you use AI.
中文: RT @thedevranjan:我排在 @conso_xyz 的 Zap 反应堆早期访问站上，排在第10963位 ⚡ 由 @SuiNetwork 和 @WalrusProtocol 提供支持 只有15000个点，Go声称属于你: 康索是您的人工智能使用私人追踪器，不会将您排除在电脑之外，并会停留在浏览器内。 真正洞察你的提示、平台、会话和习惯。分析你的活动，并改进你使用人工智能的方式。
#166 🔗 原推
2026-09-02 14:50:36
Walrus
RT @riky70229: Just started exploring @conso_xyz ⚡️ Conso is building a next-gen consumer reputation layer powered by @WalrusProtocol & @SuiNetwork. Really like the idea of turning our activity and reputation into something meaningful across the AI ecosystem. Excited to see 👀 https://twitter.com/riky70229/status/2094768155587756035/photo/1
中文: RT @riky70229:刚刚开始探索 @conso_xyz ⚡️ 康索正在打造由@WalrusProtocol & @SuiNetwork 提供支持的下一代消费者声誉层。 非常喜欢将我们的活动和声誉转化为人工智能生态系统中有意义的东西。 很高兴看到👀
media 0 共 1 项
#165 🔗 原推
2026-09-02 14:50:23
Walrus
RT @NBCryptodotsui: WalNotes is making note-taking truly Web3-native No passwords. No constant refreshing. Just a smooth, real-time workspace powered by Sui Wallet & Walrus. 📝 Live multi-tenant notes sync instantly 🔐 Sign in with your Sui Wallet 🗂️ Notes, Bookmarks & Tasks plus custom cabinets 📌 Free-form boards + Explorer directory view 🔎 Instant search + MemWal semantic recall ♻️ Trash & Archives with easy restore 🎨 11 colors, custom backgrounds & note icons 💳 Free, Premium, Founding Walrus & Memory add-on plans A simple workspace with powerful Web3 infrastructure underneath. @wal_notes × @SuiNetwork × @WalrusProtocol 🦭💙
中文: RT @NBCryptodotsui:WalNotes 正在让笔记成为真正的 Web3 原生 没有密码。没有持续的清新。只需一个由Sui Wallet & Walrus驱动的流畅、实时的工作空间。 📝 实时多租户音符即时同步 🔐 使用您的苏钱包登录 🗂️ 笔记、书签和任务以及定制橱柜 📌 自由形式板 + 资源管理器目录视图 🔎 即时搜索 + MemWal 语义召回 ♻️ 垃圾与档案，轻松恢复 🎨 11种颜色、自定义背景和附注图标 💳 免费、高级、开发海象与记忆插件套餐 一个简单的工作空间，下方配备强大的Web3基础设施。 @wal_notes × @SuiNetwork × @WalrusProtocol 🦭💙
media 0 共 1 项
#164 🔗 原推
2026-09-02 14:50:05
Walrus
RT @EdCriptoFi: AMA on @WalrusProtocol starts in 20 min. 🔗https://discord.gg/walrusprotocol Ecclusive Wal Merch ( and Memory Packs from @wal_notes ) for the best questions!
中文: RT @EdCriptoFi:@WalrusProtocol 上的 AMA 将于 20 分钟首播。 🔗 万合会海合(以及来自@wal_notes的内存包)，提供最佳问题!
#163 🔗 原推
2026-09-02 14:49:37
Walrus
RT @wal_notes: AMA on @WalrusProtocol starts in 20 min. 🔗https://discord.gg/walrusprotocol Ecclusive Wal Merch ( and Memory Packs from @wal_notes ) for the best questions!
中文: RT @wal_notes:@WalrusProtocol 上的 AMA 将于 20 分钟首播。 🔗 万合会海博物达(以及@wal_notes 提供的内存包)，提供最佳问题!
#162 🔗 原推
2026-09-02 14:49:33
Walrus
Building on Walrus? Come talk about it. Join us September 2nd at 3pm UTC for a Discord Builder AMA with @Edcriptofi, Founder of @wal_notes, a notetaking app powered by decentralized storage on Walrus. Drop your questions in our @Discord [ama-questions] channel for a chance to win exclusive Walrus merch.
#161 🔗 原推
2026-09-02 09:14:00
Walrus
RT @Philose: @WalrusProtocol @astros_ag @HumanXCo @SuiNetwork @Revolut @TheBlockCo @ThePaypers August was a WAL month 🦭
中文: RT @Philose:@WalrusProtocol @astros_ag @HumanXCo @SuiNetwork @Revolut @TheBlockCo @ThePaypers 八月是一个WAL月🦭
#160 🔗 原推
2026-09-02 04:27:10
Walrus
RT @wal_notes: Set the reminder for the Official Launch of WalNotes on Walrus discord! Exclusive @WalrusProtocol merch for the best questions! 🤯 Be there!
中文: RT @wal_notes:为WalNotes正式发布关于Walrus的不和设置提醒! 独家@WalrusProtocol merch，获取最佳问题!🤯 到那里!
#159 🔗 原推
2026-09-02 04:26:17
Walrus
RT @xpsyk0w: AI agents can execute trades. But can they trust the data they're using? @WalrusProtocol and Astros are working on that problem with WVTS — an open standard for verifiable trading data. Trades, positions, liquidations and account history stored on Walrus. Integrity verifiable through sui:native . A standardized format that other platforms can adopt. The next step for agentic finance isn't only better agents. It's giving those agents better, verifiable data.
中文: RT @xpsyk0w:人工智能代理可以执行交易。 但他们能相信他们使用的数据吗? @WalrusProtocol 和 Astros 正在利用 WVTS 处理这一问题——WVTS 是可验证交易数据的开放标准。 存储在 Walrus 上的交易、头寸、清算和账户历史。 通过 sui:native 验证完整性。 一种可采用其他平台的标准化格式。 代理金融的下一步不仅仅是更好的代理。 它为这些代理提供了更可靠、更可验证的数据。
media 0 共 1 项
#158 🔗 原推
2026-09-02 04:25:29
Walrus
RT @suidevelopers: Sui Overflow 2026 winners are in. 16 track winners across the Agentic Web, DeFi & Payments, @WalrusProtocol and @DeepBookonSui Every project below is live today ↓
中文: RT @suidevelopers:Sui Overflow 2026 获奖者将进入。 16 个跨 Agentic Web、DeFi & Payments、@WalrusProtocol 和 @DeepBookonSui 的赛道获奖者 以下每个项目今天都在直播
#157 🔗 原推
2026-09-02 04:25:21
Walrus
RT @natt_2916: @WalrusProtocol @blockticity massive flex congrats walrus this is the kind of real adoption we love to see
中文: RT @natt_2916:@WalrusProtocol @blockticity 大量灵活祝贺海象，这是我们非常希望看到的真正采用
#156 🔗 原推
2026-09-02 04:23:30
Walrus
RT @EdCriptoFi: Set reminder and be there! I'm taking notes already with @wal_notes !
中文: RT @EdCriptoFi:设置提醒并到那里! 我已经在@wal_notes上做笔记了!
#155 🔗 原推
2026-09-01 20:22:29
Walrus
RT @EdCriptoFi: @eyezenhour https://x.com/walrusprotocol/status/2094881105992397100?s=46
#154 🔗 原推
2026-09-01 20:14:28
Walrus
RT @stakecraft: @WalrusProtocol @astros_ag @HumanXCo @SuiNetwork @Revolut @TheBlockCo @ThePaypers The Verifiable Trading Standard angle is the one to watch. On-chain auditable market data that agents can actually read and verify is infrastructure most AI trading projects quietly skip, then get wrecked by bad data feeds. Astros building on this early is a smart call.
#153 🔗 原推
2026-09-01 20:14:17
Walrus
Decentralized notes, stored on Walrus. Tune in tomorrow for our next Builder AMA featuring @Edcriptofi, Founder of @wal_notes. We'll be diving into how they use Walrus for storage and answering questions live. We're giving away exclusive merch for the best questions shared in our @discord [ama-questions] channel! 📅 Sept 2 @ 3 PM UTC 📍 Walrus Discord: https://discord.com/invite/walrusprotocol You'll want to bookmark this one.
中文: 存储在海象上的去中心化笔记。 明天将收听我们的下届Builder AMA，邀请@dedcriptofi，@wal_notes创始人。我们将深入探讨他们如何使用海象进行存储并实时回答问题。 我们正在赠送独家商品，以获取我们@discord [ama-questions]频道中分享的最佳问题! 📅 9月2日 @ UTC 下午3点 📍 海象碟: 你会想为这个书签。
media 0 共 1 项
#152 🔗 原推
2026-09-01 20:12:31
Walrus
RT @FavouritonX: @WalrusProtocol @astros_ag @HumanXCo @SuiNetwork @Revolut @TheBlockCo @ThePaypers We achieved so much
中文: RT @FavouritonX:@WalrusProtocol @astros_ag @HumanXCo @SuiNetwork @Revolut @TheBlockCo @ThePaypers 我们取得了如此大的成就
#151 🔗 原推
2026-09-01 19:55:48
Walrus
Never miss a Walrus Update. Sign up for the newsletter: https://walrus.xyz/newsletter/
中文: 永远不要错过海象更新。注册新闻简报:
#150 🔗 原推
2026-09-01 19:53:25
Walrus
August brought verifiable AI trading, new Walrus Memory tutorials, and $WAL to 80M+ Revolut users. Here is everything that shipped: ⬛️ Launched the Walrus Verifiable Trading Standard with @Astros_ag to make market data machine-readable and auditable. ⬛️ Shipped three new tutorials for adding portable Walrus Memory across TypeScript SDK, Claude Code, and Claude Desktop. ⬛️ Published a 4-part deep dive breaking down agentic memory and context engineering. ⬛️ $WAL went live on @Revolut, expanding access to 80M+ users across 40+ countries. ⬛️ @KostasCryptos broke down verifiable infra on @TheBlockCo's Starting Block podcast and @ThePaypers. ⬛️ Rebuilt the Docs with product-first navigation across Walrus, Walrus Memory, and Walrus Sites. ⬛️ Catch us next month at @HumanXCo in Amsterdam, and grab your tickets for Sui Basecamp in Singapore (Oct 7–8). Need links to the events? Let us know in the comments and we can reshare them.
media 0 共 1 项
#149 🔗 原推
2026-09-01 19:51:54
Walrus
RT @triremetrading: @WalrusProtocol @blockticity $7.7B+ in real-world assets moving toward verifiable onchain records is a strong use case.
#148 🔗 原推
2026-09-01 14:58:54
Walrus
RT @wal_notes: Something is coming! Stay tuned on @WalrusProtocol discord: https://discord.com/invite/walrusprotocol https://twitter.com/wal_notes/status/2094779556724904049/photo/1
中文: RT @wal_notes:有事要来了! 敬请关注@WalrusProtocol的不和:
media 0 共 1 项
#147 🔗 原推
2026-09-01 14:20:32
Walrus
Onchain verification + private access control = real-world scale. Huge milestone with @blockticity bringing $7.7B+ in trade records to Walrus.
#146 🔗 原推
2026-09-01 14:20:08
Walrus
RT @FavouritonX: Good News: @blockticity is choosing @WalrusProtocol as it’s decentralized storage. https://twitter.com/FavouritonX/status/2094789795402723744/photo/1
中文: RT @FavouritonX:好消息:@blockticity 正在选择 @WalrusProtocol，因为它具有去中心化存储的存储空间。
media 0 共 1 项
#145 🔗 原推
2026-09-01 14:18:01
Walrus
RT @ZanzibarVenturz: GM! Proud of our team working with @WalrusProtocol and the team at @Mysten_Labs … LFG!!
中文: RT @ZanzibarVenturz:总经理! 为我们的团队与@WalrusProtocol以及@Mysten_Labs的团队合作感到自豪......LFG!!
#144 🔗 原推
2026-09-01 14:17:52
Walrus
Around 4B paper documents circulate through the $25T global trade market, making document fraud a massive, systemic risk. @blockticity is changing that by bringing over 1M authenticated records ($7.7B+ in real-world assets) onto Walrus. By mirroring 15TB+ of trade evidence across five commercial verticals, Blockticity gives partners public, cryptographic proof of origin while keeping sensitive commercial data protected. Read the full announcement: https://blog.walrus.xyz/blockticity-brings-7b-in-authenticated-trade-records-to-walrus-verifiable-data-platform/
中文: 大约4B的纸质文件在25英镑的全球贸易市场中流传，使文件欺诈成为巨大的系统性风险。 @blockticity 正在通过将超过 100 万条经过验证的记录(真实资产为 7.7 亿美元以上)引入 Walrus 来改变这一变化。 通过在五个商业垂直领域反映15TB以上的贸易证据，Blockticity为合作伙伴提供公开的加密原产地证明，同时保护敏感的商业数据。 阅读完整公告:
media 0 共 1 项
#143 🔗 原推
2026-09-01 14:06:19
Walrus
"The party walks into Castle Ravenhurst. Who greets them?" AI DM: Resurrects the boss defeated 3 sessions ago. See how @MrRobotJi used #WalrusMemory to build Aetheris: an engine that solves tabletop AI amnesia with deterministic vitality ledgers and SEAL encryption. Read the build on @Medium: https://t.co/9TvzvldwKr…
中文: 派对走进了拉文赫斯特城堡。谁在迎接他们? AI DM:复活时，老板在三节课前就失败了。 了解 @MrRobotJi 如何使用 #WalrusMemory 构建 Aetheris:一种通过确定性生命力账本和 SEAL 加密技术解决桌面人工智能失忆功能的引擎。 在@Medium上阅读该版本:
#142 🔗 原推
2026-08-31 22:37:00
Walrus
Catch the Walrus team on the ground for major announcements, partner showcases, and builder sessions. Grab your pass below and let's connect IRL: https://www.sui.io/basecamp
中文: 赶在地面上观看海象队，参加重大公告、合作展示和建设者会议。 在下方获取您的通行证，然后连接IRL:
#141 🔗 原推
2026-08-31 19:36:12
Walrus
RT @7Skyyy__: I'm going to Sui Basecamp during Token2049 hosted by @SuiNetwork Good News, @SuiCommunity_MY is giving out 2 tickets for Malaysian Developer's or Founders. Who's heading to Singapore this 7 - 8 October? Comment Basecamp! Build your own Cards 👇 https://cards.buildonsui.io/ #Merdeka #buildonsui @justusesplash
中文: RT @7Skyyy__:我将在 @SuiNetwork 主持的 Token2049 期间前往 Sui Basecamp 好消息:@SuiCommunity_MY 将为马来西亚开发者或创始人提供两张门票。谁将于10月7日至8日前往新加坡? 评论 Basecamp! 创建自己的卡片 👇 #默德卡 #buildonsui @justusesplash
media 0 共 1 项
#140 🔗 原推
2026-08-31 19:21:32
Walrus
RT @jubenrile00: Memory that lives in your repo, not just your machine.
中文: RT @jubenrile00:存储存储，不仅存储到您的设备，更可重复使用。
#139 🔗 原推
2026-08-31 05:45:13
Walrus
RT @Rainbowsdotsui: Congratulations to all the amazing winners Let's keep building with @WalrusProtocol!
中文: RT @Rainbowsdotsui:祝贺所有出色的获奖者 让我们继续使用 @WalrusProtocol 来构建!
#138 🔗 原推
2026-08-31 03:27:30
Walrus
RT @santiago_xvega: @WalrusProtocol @SuiHubAfrica la persistencia entre sesiones es el verdadero diferenciador, sin eso el asistente arranca en blanco cada vez y el usuario nota el vacío aunque el resto funcione bien
#137 🔗 原推
2026-08-31 03:27:19
Walrus
When an AI cannot track where its information comes from, you waste hours double-checking every answer. Verifiable memory keeps the system grounded in real data.
中文: 当人工智能无法追踪其信息的来源时，你会浪费数小时来对每一个答案进行双重检查。 可验证的内存使系统以真实数据为基础。
#136 🔗 原推
2026-08-30 15:23:00
Walrus
Incredible energy coming out of the hackathon at @SuiHubAfrica. Builders shipped everything from creative assistants to health trackers, all using Walrus Memory to keep context intact across tasks. This is what happens when you give AI a reliable memory system.
中文: 在@SuiHubAfrica的黑客马拉松中，带来了惊人的能量。 建筑商将从创意助理到健康追踪器的所有产品发货，均使用“海象记忆”(Walrus Memory)来在各任务中保持完整的上下文。 当你为人工智能提供一个可靠的内存系统时，就会发生这种情况。
#135 🔗 原推
2026-08-30 02:16:27
Walrus
RT @Smigglemedia: Great morning Frens Happy Weekend Xplorers 🥶 Mutuals pick up the call… A lot of edits to deliver but my mind is racing. How we building money or is it grass we’re touching today? Another day to be 1% better with @SuiHubAfrica Let’s build legendary w @WalrusProtocol🦭💙 https://twitter.com/Smigglemedia/status/2093632346394333539/photo/1
中文: RT @Smigglemedia:早上好，弗恩斯 周末快乐Xplorers 🥶 互接电话...... 有很多修改要进行，但我的思维却在加速。我们如何赚钱，还是今天接触的草? 再提一日，@SuiHubAfrica 将比现在好1% 让我们在@WalrusProtocol🦭💙中打造传奇。
media 0 共 1 项
#134 🔗 原推
2026-08-30 02:15:27
Walrus
RT @EdCriptoFi: Next week: @WalrusProtocol AMA on Discord! Public Launch!
中文: RT @EdCriptoFi:下周: @WalrusProtocol AMA 在 Discord 上! 公开发布!
#133 🔗 原推
2026-08-30 02:14:23
Walrus
Artists lose great work when the context behind an idea disappears. Audora is an excellent example of using Walrus Memory to save unfinished ideas so creators can pick up right where they left off. Shoutout to @Glo_rious9 for attending the IRL Walrus Sessions Hackathon with @SuiHubAfrica.
中文: 当一个想法背后的背景消失时，艺术家们就会失去大量作品。Audora 是利用 Walrus Memory 保存未完成想法的绝佳范例，让创作者能够直接选择离家出走的地方。 向@Glo_Rious9大声喊叫，以与@SuiHubAfrica一起参加IRL海象赛马会的精彩精彩。
#132 🔗 原推
2026-08-30 02:13:08
Walrus
RT @matt_kaschel: @WalrusProtocol @IBuzovskyi Nice product. And I love your design language.
#131 🔗 原推
2026-08-29 01:57:37
Walrus
RT @Community_Sui: Ready to move your AI memory wherever you want? 😍 Walrus Memory was born with the mission to solve that problem! Independent of any LLM 💧 #Sui #Walrus
🎬
视频
#130 🔗 原推
2026-08-28 15:49:54
Walrus
RT @EsTharrrrr: @Smigglemedia @SuiHubAfrica @WalrusProtocol @SuiNetwork @SuiFoundation @EmanAbio @kai_hudi @heroyangv @Bankyo @AvatarArtGames Love this recap. Offline prompt jams like this are how more people actually get their hands on Walrus memory without the usual Web3 barrier. Lagos energy looks strong
#129 🔗 原推
2026-08-28 15:49:02
Walrus
RT @Jife790: @Smigglemedia @SuiHubAfrica @WalrusProtocol @SuiNetwork @SuiFoundation @EmanAbio @kai_hudi @heroyangv @Bankyo @AvatarArtGames real builders, hands on learning, and portable memory coming together. Walrus is cooking.
#128 🔗 原推
2026-08-28 15:48:50
Walrus
RT @Veeekthorr: @Smigglemedia @SuiHubAfrica @WalrusProtocol @SuiNetwork @SuiFoundation @EmanAbio @kai_hudi @heroyangv @Bankyo @AvatarArtGames Nice recap. Lagos builders cooking with Walrus Memory
#127 🔗 原推
2026-08-28 15:48:44
Walrus
RT @Mysten_Labs: Another record loading.
#126 🔗 原推
2026-08-28 15:46:40
Walrus
RT @kimblgn: @petergyang not sure if you have tried walrus wemory yet, but your post describes the exact use case we are working on https://docs.wal.app/walrus-memory/mcp/overview
中文: RT @kimblgn:@petergyang 不确定您是否尝试过海象的骵默，但您的帖子说明了我们正在处理的具体使用案例
#125 🔗 原推
2026-08-28 15:33:32
Walrus
RT @kimblgn: This is exactly the user behavior we had in mind when we started building Walrus Memory The more context people build with AI, the less willing they will be to start from zero every time they try a new application And our goal was simple. People should own that context and decide which applications can use it
中文: RT @kimblgn:这正是我们开始构建 Walrus Memory 时想到的用户行为 人们使用人工智能构建的语境越多，每次尝试新应用程序时，他们就越不愿意从零开始 我们的目标很简单。人们应该拥有这种背景，并决定哪些应用程序可以使用它
#124 🔗 原推
2026-08-28 15:32:14
Walrus
RT @0xd34th: @Smigglemedia @SuiHubAfrica @WalrusProtocol @SuiNetwork @SuiFoundation @EmanAbio @kai_hudi @heroyangv @Bankyo @AvatarArtGames W Session need more of these
#123 🔗 原推
2026-08-28 15:30:04
Walrus
RT @Smigglemedia: @Ameenixonweb3 @SuiHubAfrica @WalrusProtocol @SuiNetwork @SuiFoundation @EmanAbio @kai_hudi @heroyangv @Bankyo @AvatarArtGames The best builders are building with @WalrusProtocol 🦭
#122 🔗 原推
2026-08-28 15:29:43
Walrus
RT @Smigglemedia: Mentioned you… Remember The Walrus Memory x AI Prompt Hackathon? Here’s a quick highlight of the session in Lagos @SuiHubAfrica - Builders learnt prompt design taking @WalrusProtocol online Prompt Jam 5 format offline - The session gave builders hands-on guidance to explore what is possible with portable agent memory. -Some builders worked together and provided solutions to real problems I think @0xd34th and @chowtato would be proud of this session The best builders are cooking with @WalrusProtocol Gm & Happy Friday Xplorers 🦭💙 Stay tuned for more…
🎬
视频
#121 🔗 原推
2026-08-28 04:22:24
Walrus
RT @SuiHubAfrica: That's a wrap on Session Lagos: Walrus Memory x AI Prompt at the Hub🎥 Today, we had a full house of builders at the SuiHub come together to learn, experiment, collaborate, and turn ideas into builds using Walrus Memory × AI prompts. A big thank you to everyone who showed up and participated! Results and top submissions dropping soon
🎬
视频
#120 🔗 原推
2026-08-28 03:10:45
Walrus
RT @gui01macedo: proud of being a part of this 🫡
中文: RT @gui01macedo:为自己参与其中而感到自豪 🫡
#119 🔗 原推
2026-08-28 03:10:27
Walrus
RT @0xMahdieh: And @WalrusProtocol can solve this
中文: RT @0xMahdieh:@WalrusProtocol 可以解决这个问题
#118 🔗 原推
2026-08-28 03:09:09
Walrus
We are bringing the future of agentic memory to the floor with major announcements, stage keynotes, and hands-on workshops. Stop by the booth to see live partner showcases and meet the Walrus team. There's still time to get your tickets: https://www.sui.io/basecamp
中文: 我们将通过重大公告、舞台主题演讲和实践研讨会，将特化记忆的未来带入现场。 在展台旁参观现场对讲，与华勒斯团队见面。 还有时间购买门票:
#117 🔗 原推
2026-08-27 20:19:04
Walrus
RT @EdCriptoFi: @WalrusProtocol protocol Stack. Unforgettable and Encypted Data.
中文: RT @EdCriptoFi:@WalrusProtocol 协议栈。 令人难忘的和被吸收的数据。
#116 🔗 原推
2026-08-27 17:33:26
Walrus
RT @eazitechh: @WalrusProtocol @SuiHubAfrica @Smigglemedia It was really an insightful session learning about Walrus Memory Can't wait to submit a memory system prompt i have built
中文: RT @eazitechh:@WalrusProtocol @SuiHubAfrica @Smigglemedia 这是一次关于 Walrus Memory 的深刻见解 迫不及待想提交我已经构建的内存系统提示
#115 🔗 原推
2026-08-27 17:31:35
Walrus
Live in Lagos @SuiHubAfrica: over 30 builders gathered in person to learn prompt design and build with Walrus Memory and context. Taking our online Prompt Jam 5 format offline, this session gives builders hands-on guidance to explore what is possible with portable agent memory. Looking forward to seeing what the teams build.
中文: 住在拉各斯 @SuiHubAfrica:30多名建筑工人亲自聚集在一起，学习使用海象记忆和背景进行快速设计和施工。 将我们的在线Prompt Jam 5格式离线，本课程为开发者提供实践指导，以探索便携式代理内存的可能。 期待看到团队的发展。
media 0 共 4 项 media 1 media 2 media 3
#114 🔗 原推
2026-08-27 17:27:00
Walrus
RT @eazitechh: Live at @SuiHubAfrica Building with Walrus Memory https://twitter.com/eazitechh/status/2092999838656872922/video/1
中文: RT @eazitechh:@SuiHubAfrica 现场直播 使用 Walrus 记忆构建
🎬
视频
media 1
#113 🔗 原推
2026-08-27 17:09:17
Walrus
RT @inkray_io: 🏆 @WalrusProtocol Sessions 6 winners have been announced! Huge congrats to everyone who wrote, published, and shared their take on AI Agents & Walrus Memory; and to everyone who shared valuable feedback and climbed the Inkray leaderboard. Think you made the list? 👀 Check the Walrus Discord to see if you won! Big congrats to all the creators who made this round worth remembering!
🎬
视频
#112 🔗 原推
2026-08-27 17:09:10
Walrus
RT @Smigglemedia: Great Morning Frens Happy Thursday Xplorers🥶 Today’s the day! 📍The Walrus Memory x AI Prompt Hackathon is live and will start by 12pm at @SuiHubAfrica Lagos. -Physical presence is required to participate. -No coding or web3 background needed. -Just come with your laptop, an idea, and your curiosity for a chance at $300 WAL Let’s build legendary together💙
media 0 共 2 项
🎬
视频
#111 🔗 原推
2026-08-27 14:45:02
Walrus
RT @suidevelopers: Walrus ( @WalrusProtocol ) 1. Kraterion: S3-compatible storage you actually own, with agents on top 2. Deploy by @suize_io: push a site to the decentralized web in one call 3. @CarrySui: agent memory that proves what it used 4. Praxis: simulate every agent spend before it signs
中文: RT @suidevelopers:海象(@WalrusProtocol) 1。Kraterion:您实际拥有的S3兼容存储，顶部为代理 2。由 @suize_io 部署:一次调用将网站推送到去中心化网络 3。@CarrySui:代理内存，证明了它所使用的内容 4.实践:在每个代理出现迹象之前，请先模拟其支出
#110 🔗 原推
2026-08-27 14:36:29
Walrus
RT @SuiNetwork_JP: Sui Overflow 2026、16のトラック受賞プロジェクトが決定。 Agentic Web、DeFi & Payments、@WalrusProtocol 、@DeepBookonSui の4トラックに加え、10組の学生チームがUniversity Awardを受賞しました。 全26組の受賞プロジェクト・チームはこちら↓ https://twitter.com/SuiNetwork_JP/status/2092802340881654082/photo/1
media 0 共 1 项
#109 🔗 原推
2026-08-27 14:35:31
Walrus
RT @7Skyyy__: Splash won the DeFi & Payments track at Sui Overflow 2026 🏆 — Top 4. What we submitted: B2B settlement for Southeast Asia, built on @SuiNetwork and designed to scale globally. USD/USDC in, fiat out, with cryptographic receipts anchored on-chain. Omnibus treasury loading. The chain is the settlement layer, not the product. Businesses never see or touch a token, no gas, no seed phrases, no crypto UX. Just payments that prove themselves. Built with Move, @WalrusProtocol, Seal, and zkLogin. Settle → Save → Supply. Payments are the wedge; this is loop one. Phase 1 of 5 : long road ahead, and the next mile is turning testnet proof into live corridors. Congrats to @taliseio and every team that shipped — and thank you to the judges, the @SuiFoundation , @suidevelopers, @SuiCommunity, and @SuiCommunity_MY. This ecosystem builds. Thank you SEAblings @SuiCommunity_ID & @SuiCommunity_PH Handsome Engineer : @sebestdebest
中文: RT @7Skyyy__:Splash 赢得了 Sui Overflow 2026 的 DeFi & Payments 曲目，该 🏆 已 获得第 4 名。 我们提交的:以@SuiNetwork为建，旨在扩展全球规模的东南亚B2B结算。美元/美元兑美元，用法币出价，加密凭证固定在链上。全库装。 链条是沉降层，不是产品。企业从不看到或触碰代币，没有燃气，没有种子短语，没有加密用户体验。只是证明自己的支付。 使用 Move、@WalrusProtocol、Seal 和 zkLogin 构建。 结算 → 保存 → 供应。支付是楔子;这是循环一。 第一阶段(5期):前方漫长的道路，下一英里正将测试网证明变成实时通道。 祝贺@taliseio以及所有发货团队——感谢评委们、@SuiFoundation、@suidevelopers、@SuiCommunity 和 @SuiCommunity_MY。这个生态系统正在构建。 谢谢 SEABLBYS @SuiCommunity_ID 和 @SuiCommunity_PH 帅工工程师 :@sebestdebest
media 0 共 1 项
#108 🔗 原推
2026-08-27 14:33:08
Walrus
RT @CarrySui: We're excited to share that Carry took 3rd place in the @WalrusProtocol Track at Sui Overflow 2026! Carry is a proof layer for AI agent memory, making memory-based AI answers provable and verifiable. A huge thank you to the judges, the organizers, and everyone who supported us throughout the hackathon. This is just the beginning, we've got more exciting updates to share in the coming days. Stay tuned! https://usecarry.xyz/
media 0 共 1 项
#107 🔗 原推
2026-08-27 14:32:15
Walrus
RT @dreyethh: first time building on walrus protocol btw it’s an honor 🙏❤️..let’s do more
#106 🔗 原推
2026-08-27 14:26:44
Walrus
Moving beyond basic agent demos requires infrastructure that can handle real value and state onchain. Congrats to the @Talus_Labs team on launching v2.0 on @SuiNetwork.
中文: 超越基本代理演示需要能够处理实际价值和状态链的基础设施。 祝贺 @Talus_Labs 团队在 @SuiNetwork 上推出 v2.0。
#105 🔗 原推
2026-08-27 01:27:00
Walrus
Congrats to all the winners of the Sui Overflow hackathon! Incredible work across the board. The winners of this year's Walrus track focused on rethinking how agentic systems (including memory) are built, and each won their share of the $70k prize pool: 🥇 – Kraterion: S3-compatible object storage where users retain full ownership, featuring an integrated AI runtime for tamper-evident agent runs. 🥈 – Deploy by Suize: A deployment tool that pushes websites to the decentralized web in a single call, hosting files on Walrus and verifying hashes on-chain. 🥉– Carry: A memory system for AI agents that forces permission checks and cryptographic proof before an agent can read stored context. 4th place – Praxis: Security middleware that dry-runs every agent transaction, risk-scores the spend, and logs decision receipts directly to Walrus.
#104 🔗 原推
2026-08-27 01:04:04
Walrus
RT @SuiHubAfrica: Tomorrow’s the day! The Walrus Memory x AI Prompt Hackathon will start tomorrow at SuiHub Lagos. 📍 Physical presence is required to participate. Come with your laptop, an idea, and your curiosity. Let’s build!
中文: RT @SuiHubAfrica:明天是这一天! “海象记忆” x 人工智能提示黑客马拉松将于明天在拉各斯苏伊伯纳开赛。 📍 参加需要实际参与。 带着你的笔记本电脑，一个想法，以及你的好奇心。让我们来建设!
#103 🔗 原推
2026-08-27 00:19:14
Walrus
RT @cryptosetters: @WalrusProtocol Big milestone for Walrus 🐋🔥 Persistent memory is a powerful primitive, and seeing developers turn it into real ideas is even more exciting. Congrats to everyone who participated! 👏
中文: RT @cryptosetters:@WalrusProtocol 对 Walrus 而言，持久内存是一种强大的原始功能，而开发者将其转化为真实想法，更加令人兴奋。向所有参与的人们点好!👏
#102 🔗 原推
2026-08-26 19:44:29
Walrus
RT @cryptosetters: @WalrusProtocol @dexarxbt Milestone.... This is where persistent memory gets really interesting 🔥 It’s not just about remembering data, but understanding the “why” behind decisions across sessions. Great implementation by @dexarxbt and ScholarFlow 👏🐋
中文: RT @cryptosetters:@WalrusProtocol @dexarxbt 里程碑...... 在这里，持久性记忆变得非常有趣🔥，它不仅在于记住数据，而在于理解跨会话决策背后的“原因”。优秀的实现者:@dexarxbt 和 ScholarFlow 👏🐋
#101 🔗 原推
2026-08-26 19:44:23
Walrus
RT @SuiCommunity: Paris. Dubai. Now Singapore. Our favorite part of Basecamp is meeting the people you have only ever seen in a group chat. Six weeks out. Who is flying in? 🇸🇬
中文: RT @SuiCommunity:巴黎。迪拜。现在是新加坡。 我们在Basecamp中最喜欢的部分是与那些你只在群聊中见过的人见面。 六周后。谁在飞?🇸🇬
#100 🔗 原推
2026-08-26 19:40:55
Walrus
RT @SuiCommunity: . @SuiNetwork is Basecamp maxxing! Who all are joining the intern in Singapore? https://twitter.com/SuiCommunity/status/2092299962344460557/photo/1
media 0 共 1 项
#99 🔗 原推
2026-08-26 05:06:28
Walrus
RT @eazitechh: I built exam mistake memory prompt, exam mistake memory was built to help students stop wasting time on what they already know. I'm a Pharm.D student. Every time I studied with an AI, the session died when I closed the tab. Next day it would re-explain things I already understood and had zero memory of the questions I kept failing. That's the only data that actually matters before an exam, and it was disappearing every single night. Then I found Walrus Memory. Memory that lives on Walrus and follows your wallet instead of the app. That was the fix. So I wrote a prompt around it for the Prompt Jam and it ended up winning. Here's how to use it: 1. Install the MemWal MCP server in your client. Claude Code, Codex, Cursor all work the same way: npx -y @mysten-incubation/memwal-mcp 2. Run memwal_login. It opens a browser wallet sign in, link expires in 5 minutes so approve it fast. You never touch a private key. 3. Copy the full prompt from the repo and paste it into your CLAUDE.md or your agent's system prompt. 4. Fill the config block at the top. Your exam, your subjects, one short tag per subject. 5. Start answering past questions. Get things wrong. Every wrong answer writes a blob to Walrus Mainnet on its own, no command needed. 6. Close everything, come back tomorrow, say "prep me". It opens with your five worst topics ranked by how often and how badly you missed them. Prompt: https://github.com/EAZITECH1/exam-mistake-memory Demo video: https://youtu.be/o_B-LKcSS9g
media 0 共 1 项
#98 🔗 原推
2026-08-26 05:06:02
Walrus
RT @eazitechh: @WalrusProtocol Thanks for the Spotlight Walrus:
#97 🔗 原推
2026-08-26 05:05:51
Walrus
RT @kimblgn: AI is turning intelligence from a scarce capacity bound to humans into something that can be produced, replicated and deployed at scale But intelligence without context is generic. What makes it useful is everything it understands about our history, preferences and intent The big opportunity here is to make any model or application more useful to users by bringing the right context into every situation That is the utility Walrus Memory is designed to unlock
#96 🔗 原推
2026-08-25 16:45:09
Walrus
Context engineering decides what a model sees at each step: instructions, history, tools, and recalled memory. But temporary context cannot hold state across runtimes. @JessieWritesx, Tech Lead Manager at @Mysten_Labs, walks through where context engineering stops and durable memory takes over. Full write-up: https://blog.walrus.xyz/what-is-context-engineering/
media 0 共 1 项
#95 🔗 原推
2026-08-25 14:24:52
Walrus
That's a wrap on the Walrus Memory Prompt Jam. Thank you to every developer, writer, and builder who submitted prompts and showcased the power of persistent memory. Explore all project submissions and open-source prompts in the full showcase on DeepSurge: https://www.deepsurge.xyz/hackathons/f313beb4-290d-46d9-ac73-3e216fdba8d1
#94 🔗 原推
2026-08-25 00:00:59
Walrus
RT @AA_RonOnChain: Built a multiverse storyteller for my kids for Walrus Session 7, evolving Continuity Keeper with Scoped Canon. One character. Many realities. Shared identity, different histories. The interesting part was figuring out what memory should do when two conflicting memories can both be true, just in different worlds. Article: https://medium.com/@AA_RonOnChain/how-i-evolved-continuity-keeper-into-a-multiverse-storyteller-for-my-kids-fbab6be74938 Repo: https://github.com/ajw72787/walrus-session7-scoped-canon @WalrusProtocol #WalrusMemory
media 0 共 1 项
#93 🔗 原推
2026-08-24 23:59:40
Walrus
RT @sadboy_042: @WalrusProtocol I gave Claude Code a memory on Walrus Mainnet and built with it for 3 days. It wrote 26 records. I tested the original prompt against my evolution: v1 serves a stale decision as current, v2 doesn't. Story: https://medium.com/@makindedaniel45/my-claude-code-sessions-kept-forgetting-everything-buildmem-v2-on-walrus-memory-14331ea069c2 @WalrusProtocol #Walrus #WalrusMemory https://twitter.com/sadboy_042/status/2091735432744595637/video/1
media 0 共 2 项
🎬
视频
#92 🔗 原推
2026-08-24 23:59:37
Walrus
RT @sandman_sh: Tired of re-explaining your code and stack every time you switch between Claude Code and Cursor? ⚡ For @WalrusProtocol Session 7: Prompt Evolution, I built Continuum 2.0 on #WalrusMemory! What changed: 🔹 57.7% startup token reduction (1-Shot Pointer Router) 🔹 [pref_v2] Invalidation Graph to stop old & new rules clashing 🔹 Cross-Model AST & Git-Delta State Normalizer for seamless IDE handoffs 🔹 Verified on Walrus Mainnet Read the full story: https://medium.com/@sandman.sh/stop-losing-your-mind-across-ai-tools-how-continuum-2-0-d77fe2d2d579?postPublishedType=initial
#91 🔗 原推
2026-08-24 23:58:57
Walrus
RT @0xanjalii: Excited to share Markov Engine 2.0 for @WalrusProtocol Session 7! 🧠⚡ We evolved #WalrusMemory from reactive search to predictive state transitions: 📖 Article: https://medium.com/@0xvampire/i-stopped-asking-my-ai-to-search-how-markov-chains-on-walrus-made-my-coding-agent-predict-the-4e94f413e2de?postPublishedType=initial https://twitter.com/0xanjalii/status/2091842861394464994/photo/1
media 0 共 1 项
#90 🔗 原推
2026-08-24 23:58:42
Walrus
RT @Eminent4k: I evolved the Markov prompt after testing its handoff workflow and identifying ways to make cross-agent continuity more reliable See what I changed in Markov V2: Github: https://github.com/eminentcodes/markov-v2.git Medium : https://medium.com/@greatesegbue/when-an-ai-agent-forgets-the-work-the-next-session-starts-from-zero-a2f6339cff62?sharedUserId=greatesegbue @WalrusProtocol #WalrusMemory https://twitter.com/Eminent4k/status/2091893001068941792/photo/1
media 0 共 1 项
#89 🔗 原推
2026-08-24 23:58:32
Walrus
RT @ConorBronsdon: + thank you to @chain_ofthought's presenting sponsors for enabling us to do this episode: - @WalrusProtocol: Walrus Memory provides portable, verifiable memory that carries context across agents and apps: https://walrus.xyz/cot - @SvixHQ: Reliable, webhook infra for startups and the Fortune 500 - qualified startups get $12k or more in credits: https://link.svix.com/cot
media 0 共 1 项
#88 🔗 原推
2026-08-24 23:46:20
Walrus
ICYMI: $WAL is now listed on @Revolut. Expanding access across Revolut's 80M+ userbase in 40+ countries.
#87 🔗 原推
2026-08-24 18:42:00
Walrus
RT @Philose: Portable, verifiable memory is the turning point #Walrus
#86 🔗 原推
2026-08-24 18:40:23
Walrus
RT @eazitechh: Memory that can't leave the app is memory you rebuild every session. That's the real AI problem. Not intelligence, not context length. Every quarter: smarter models, longer context windows, better benchmarks. Every Monday: agents remember nothing from Friday. The industry keeps optimizing the wrong variable. Email had this exact disease before IMAP. Your archive died with your mail client, and switching meant losing everything. Then the store moved out from under the client, and clients became interchangeable. Agent memory is stuck in the pre-IMAP era. Six months of context your agent learned belongs to the platform, not you. Walrus Memory launched by @WalrusProtocol is the inversion: memory keyed to your wallet, not the app. I wrote a full breakdown of this on Medium: https://medium.com/@ajayiisrael523/memory-you-cant-move-is-memory-you-keep-rebuilding-b155b3e1ea89 @inkray_io
media 0 共 1 项
#85 🔗 原推
2026-08-24 17:35:40
Walrus
RT @Philose: @WalrusProtocol @kostascrypto @gazza_jenks @TheBlockCo Portable, verifiable memory is the turning point
#84 🔗 原推
2026-08-24 17:35:23
Walrus
RT @alexbelij: A stored failure that cannot refuse an action is only a note. So each failure record carries its own clearance condition: the deploy stays blocked until a reviewed verifier passes on that commit. Demo: https://failure-to-gate.vercel.app/ https://alexbelij.medium.com/a-remembered-failure-should-be-able-to-stop-you-d913779dbcf6 #WalrusMemory @WalrusProtocol
#83 🔗 原推
2026-08-24 17:28:05
Walrus
RT @FavouritonX: @WalrusProtocol @kostascrypto @gazza_jenks @TheBlockCo AI is only as powerful as the memory it can carry with it. Portable AI memory will be the next major piece of the puzzle.
#82 🔗 原推
2026-08-24 17:23:26
Walrus
RT @KrugTriumph: New post: Remembering that a student got a question wrong is easy. Knowing whether it breaks the next topic is not. Transfer Engine stores mistakes with their concept edges and gates the next lesson on all of them. @WalrusProtocol #WalrusMemory https://medium.com/@triumphkrug/the-lesson-that-was-right-and-still-gave-me-the-wrong-answer-2491d3cae0d8
#81 🔗 原推
2026-08-24 17:23:19
Walrus
RT @IOxOOOxOOO: Canon breaks when two true facts are written months apart and never reconciled. @WalrusProtocol #WalrusMemory https://medium.com/@alfakalor.com/what-is-true-now-reading-an-append-only-memory-as-a-ledger-1df06bbd1a47
#80 🔗 原推
2026-08-24 17:23:18
Walrus
RT @eazitechh: I corrected one fact in Claude Code. Codex enforced it minutes later, different vendor, no shared context. That's Continuity Keeper: a memory prompt giving an AI co-writer a canon stored on Walrus, checked against every draft. I calibrated it: https://github.com/EAZITECH1/continuity-keeper-calibrated/blob/main/prompt/continuity-keeper-calibrated.md Medium Article: https://medium.com/@ajayiisrael523/every-new-ai-session-started-from-zero-until-i-fixed-it-with-walrus-memory-784e68df004b #Walrus
media 0 共 2 项 media 1
#79 🔗 原推
2026-08-24 17:01:40
Walrus
Moving your AI memory has never been more necessary. @kostascrypto @gazza_jenks on @theblockco. https://twitter.com/WalrusProtocol/status/2091888546609262911/video/1
🎬
视频
#78 🔗 原推
2026-08-24 14:01:09
Walrus
RT @oyetoludan01: EXAM MISTAKE MEMORY V2 My improvement of an exam prep prompt built by @EAZITECH1, running on Walrus Memory from @WalrusProtocol. The original had one major blind spot. It only stored wrong answers. V2 adds confidence tracking, so it captures guesses, uncertainty, and confident mistakes. It also replaces impossible updates with append only supersession, reports failed writes, and adds a three day minimum for mastery. Then the system caught a deeper pattern in my own performance. I thought I had a definition problem, but after 16/18 it showed my mistakes were actually concentrated in the final third of longer sessions. The old conclusion was not deleted. It was superseded, preserved onchain, and still verifiable. 5 sessions. 2 subjects. 55 questions. ~40 memories on mainnet. Full Medium write up: https://medium.com/@oyetoludan/being-wrong-is-not-one-thing-7c26373904d0 Prompt and evidence: https://github.com/Tolex081/exam-mistake-memory-v2 Youtube video: https://youtu.be/zgCN2TLD7fY #WalrusMemory @WalrusProtocol
media 0 共 4 项 media 1 media 2 media 3
#77 🔗 原推
2026-08-24 00:39:27
Walrus
RT @bluebloodhakani: Evolved BuildMEM Agent for @WalrusProtocol 's Prompt Evolution hackathon Found a silent bug where the agent's memory tag was a guess that drifted across sessions, plus a recall filter that was reading the tool's output backwards. 12 real memories on Walrus Mainnet prove the fix. Medium: https://medium.com/@bluebloodhakani/buildmem-agent-project-identity-resolution-recall-relevance-72530465be7c?sharedUserId=bluebloodhakani Github: https://github.com/akaniooh/buildmem-agent/tree/prompt-evolution/project-identity-and-recall-relevance #Walrus #WalrusMemory
🎬
视频
media 1
#76 🔗 原推
2026-08-24 00:39:14
Walrus
RT @GeiserJoe2: Just finished my Walrus Session 7 entry. I took the original D&D Campaign Memory prompt and turned it into a real continuity engine: • BOOT ritual that recalls the campaign every new session • Contradiction-guard that stops dead NPCs from walking • Append-only supersede so current truth wins • Clean session handoffs Ran it for real on Mainnet (namespace session7-dnd-makabeez). Session 2 correctly remembered everything. When I deliberately broke canon, it caught the conflict. When Gorruk died, it wrote a superseding record. Repo: https://github.com/Makabeez/session7-dnd-campaign-memory Article: https://medium.com/@tonystarks1220/how-i-stopped-my-d-d-campaign-from-contradicting-itself-across-sessions-with-walrus-memory-b966649d4070?sharedUserId=tonystarks1220 @WalrusProtocol #WalrusMemory
media 0 共 1 项
#75 🔗 原推
2026-08-24 00:39:03
Walrus
RT @eazitechh: @doncurrent @WalrusProtocol @Osazedotsui Exactly bro, exactly why I built it. To solve AI agent amnesia, especially with the rising use of AI among students for studying and learning. The goal is to make AI memory more persistent, portable, and useful across different study sessions.
#74 🔗 原推
2026-08-24 00:38:38
Walrus
RT @cryptosetters: @WalrusProtocol @0xanjalii Walrus keeps proving that decentralized storage can be more than just infrastructure. 🐋💾 Building real products, real use cases, and a stronger decentralized future. My favorite project for a reason. 🐋💙
#73 🔗 原推
2026-08-24 00:38:26
Walrus
RT @0xcollinxweb3: My Claude credits ran out mid bug hunt. I didn't lose the trail I recalled it on Grok and kept going, word for word, zero re-explaining. 👇 #WalrusMemory https://medium.com/@collinxweb3/free-tier-limits-no-longer-kill-my-debugging-sessions-heres-the-memory-capsule-i-use-f091b0fe358c https://twitter.com/0xcollinxweb3/status/2091239221423960210/photo/1
media 0 共 1 项
#72 🔗 原推
2026-08-24 00:36:23
Walrus
There are 24 hours left in the Walrus Memory Prompt Jam. Check out all 6 featured prompt categories, test the open-source prompts, and explore the projects on DeepSurge before submissions close tomorrow. Explore the showcase: https://www.deepsurge.xyz/hackathons/f313beb4-290d-46d9-ac73-3e216fdba8d1
#71 🔗 原推
2026-08-23 11:22:00
Walrus
Studying with an AI agent gets better when it remembers your past mistakes across course namespaces instead of resetting context. @Osazedotsui tested this live during real exam prep using Walrus Memory. Worth a read.
#70 🔗 原推
2026-08-22 23:09:00
Walrus
Most agent memory stores what changed, tracking why decisions were made across sessions is where persistent state matters. Clever implementation by @dexarxbt using BuildMEM with ScholarFlow to keep context coherent across long runs. Full demo and walkthrough below:
#69 🔗 原推
2026-08-22 19:47:00
Walrus
View on DeepSurge: https://www.deepsurge.xyz/projects/900043a3-dd2c-46bc-9bd7-b4e921464374
#68 🔗 原推
2026-08-22 16:21:18
Walrus
RT @DLougaris: Excited to share that I'll be a speaker at Sui Basecamp 2026 in Singapore (Oct 7-8) 💧 I'll be talking about @WalrusProtocol , as well as demoing Walrus Memory, which gives AI agents persistent, portable memory. Would love to see you there! https://www.sui.io/basecamp
#67 🔗 原推
2026-08-22 16:17:53
Walrus
For our final Walrus Prompt Jam community spotlight, we are highlighting a standout project for tabletop RPGs: D&D Campaign Memory by @0xanjalii on Github. Long-running tabletop campaigns generate hundreds of NPCs, quest lines, locations, and decisions that are tough to track manually. D&D Campaign Memory acts as a persistent co-DM, organizing campaign lore, session notes, and world state across months of gameplay. Explore the prompt: https://github.com/0xanjalii/Campaign-Vault/blob/main/dnd-dm-assistant.md
media 0 共 1 项
#66 🔗 原推
2026-08-22 15:36:00
Walrus
RT @golycuotas: @WalrusProtocol la primera vez que vi un agente con memoria persistente fue en una demo cerrada y lo único que pensé fue "esto por fin no se olvida de lo que le pedí hace 5 minutos" 🧠
#65 🔗 原推
2026-08-22 14:37:26
Walrus
See what portable agent memory looks like in production. Catch Dio from the Walrus Builder Growth team at Sui Basecamp in Singapore (Oct 7–8) for a live demo of Walrus Memory. Get your Sui Basecamp tickets: https://www.sui.io/basecamp
#64 🔗 原推
2026-08-22 11:44:00
Walrus
View on DeepSurge: https://www.deepsurge.xyz/projects/21b0797b-32da-493d-a5b7-e8e927a99746
#63 🔗 原推
2026-08-21 20:02:30
Walrus
Most exam prep tools re-test material students already understand. Today's Prompt Jam spotlight: Exam Mistake Memory by @eazitechh, (@/eazitech1) on Github. Exam Mistake Memory tracks errors, pinpoints specific knowledge gaps, and generates targeted practice sessions based on past attempts. Explore the prompt: https://github.com/EAZITECH1/exam-mistake-memory/blob/main/prompts/exam-mistake-memory.md
media 0 共 1 项
#62 🔗 原推
2026-08-21 20:02:14
Walrus
@kostascrypto sat down with @gazza_jenks on @TheBlockCo's Starting Block podcast to discuss AI threat models, data provenance, and where Walrus fits in. Watch the full interview: https://x.com/i/broadcasts/1yGBePVXNPkKN?s=20
#61 🔗 原推
2026-08-21 18:04:00
Walrus
RT @EmanAbio: We cracked something nobody else has. How to let humans and AI agents move real money into DeFi without losing the trust part. This is a tech breakthrough. Coming to #SuiBasecamp.
#60 🔗 原推
2026-08-21 18:03:38
Walrus
RT @SuiNetwork: Something big is coming. See it first at Sui Basecamp. https://sui.io/basecamp
#59 🔗 原推
2026-08-21 18:03:33
Walrus
RT @richie_binance: There’s an ongoing 2-week online hackathon focused on improving AI memory prompts with @WalrusProtocol Memory. No coding required. Just pick a prompt, test it, improve it, and share what you learned. check it out 👇 https://www.reddit.com/r/hackathon/s/11EJaVFhdd
#58 🔗 原推
2026-08-21 16:59:13
Walrus
RT @cryptosetters: @WalrusProtocol @github This is the kind of AI tool developers actually need. 🧠 Saving decisions, errors, and session progress means less time repeating old work and more time building. Walrus + AI is getting interesting. 🐋 https://twitter.com/cryptosetters/status/2090623127881683276/photo/1
media 0 共 1 项
#57 🔗 原推
2026-08-21 16:58:17
Walrus
RT @King_Breezyy: @Philose @WalrusProtocol Walrus is cooking hard 🦭 200+ projects + 686TB in just over a year is no joke. Storage layer of the future loading..
#56 🔗 原推
2026-08-21 16:22:27
Walrus
RT @Philose: -200+ projects -686TB of data stored -AI agents, and builders all on the same storage layer Walrus launched on mainnet in March 2025 and the ecosystem built around it in just over a year is worth highlighting Here is what is actually being built on #Walrus right now 🧵🔻 https://twitter.com/Philose/status/2090698811551158698/photo/1
media 0 共 1 项
#55 🔗 原推
2026-08-21 16:22:20
Walrus
RT @TheBlockCo: Breaking Down AI’s Threat to Cryptography /w Kostas Chalkias https://x.com/i/broadcasts/1yGBePVXNPkKN
#54 🔗 原推
2026-08-21 14:44:12
Walrus
RT @DLougaris: You are just a few lines of code away from having an agent with persistent, portable memory, give it a shot!🤖
#53 🔗 原推
2026-08-21 14:39:14
Walrus
RT @DforDouble: @Philose @WalrusProtocol Walrus has to be one of the best projects on Sui
#52 🔗 原推
2026-08-21 14:39:09
Walrus
RT @gazza_jenks: TODAY ON THE STARTING BLOCK: We've got @Mysten_Labs co-founder @kostascrypto in the Hot Seat for a deep conversation on AI, cryptography. We'll also get a pulse check on @WalrusProtocol @SuiNetwork. Join us live on @TheBlockCo https://x.com/i/broadcasts/1yGBePVXNPkKN?s=20
#51 🔗 原推
2026-08-21 14:33:37
Walrus
RT @cryptosetters: @WalrusProtocol @astros_ag This is actually interesting 👀 Turning raw onchain activity into a clear Trading DNA profile is a cool use case. Walrus + Astros is a strong combo. 🐋🧬 #Walrus #Sui #Base #Onchain
#50 🔗 原推
2026-08-21 14:29:12
Walrus
Single-key authentication gives an agent full authority and zero room for error. Once that key is compromised, there is no second gate. @Kostascrypto breaks down why multi-signature verification is becoming non-negotiable as autonomous workflows scale. https://twitter.com/WalrusProtocol/status/2090788509292634187/video/1
🎬
视频
#49 🔗 原推
2026-08-21 13:10:00
Walrus
"A memory system that lies confidently is worse than one that crashes." Thanks to @OlabanjiOlalek2 for submitting this breakdown on fixing deduplication gaps, prompt injection, and state continuity using Walrus Memory. Worth a full read for anyone building long-term agent state.
#48 🔗 原推
2026-08-20 21:42:00
Walrus
Walrus' Trading Personality is: Never Forgets. Generate your Trading DNA Card with Astros Scan, and share it on X with @astros_ag. https://twitter.com/WalrusProtocol/status/2090553355018064252/photo/1
media 0 共 1 项
#47 🔗 原推
2026-08-20 21:35:35
Walrus
RT @SuiCommunity: Sui’s trading personality is ‘Narrative Gardener’. What’s yours? 👀 https://twitter.com/SuiCommunity/status/2090517239980335454/photo/1
media 0 共 1 项
#46 🔗 原推
2026-08-20 21:05:21
Walrus
Permanent, verifiable trading data arrives on Sui. @Astros_AG launched Astros Scan, built on Walrus to transform raw market activity into searchable, open intelligence. Want to win some Astros Points? Generate your AI Trading DNA profile from your onchain history, share your card with a one-liner, and tag @Astros_AG.
#45 🔗 原推
2026-08-20 18:47:23
Walrus
Learn more about the project on DeepSurge: https://www.deepsurge.xyz/projects/cda00470-30ad-415c-8084-305a8ca2a3c8
#44 🔗 原推
2026-08-20 18:37:40
Walrus
Today we are highlighting a project for developer productivity: BuildMEM Agent by @/Olalekan2345 on @github. Debugging notes, architecture decisions, and deployment fixes vanish the moment a session ends. BuildMEM Agent captures decisions, errors, and session progress across hackathons and long-term builds so you never fix the same bug twice. Who is tackling dev productivity or building for the prompt jam right now? Tell us about your project. Explore the prompt: https://github.com/Olalekan2345/buildmem-agent/blob/main/CLAUDE.md
media 0 共 1 项
#43 🔗 原推
2026-08-20 18:37:23
Walrus
RT @0xbeepit: The next $5 trillion in transactions won't be human. It will come from agents analyzing, allocating, paying, and trading capital autonomously. That's Machine GDP in motion. Our Co-Founder & CEO @0xvati is heading to Sui Basecamp 2026 to talk about the Agentic Economy we're already building toward on @SuiNetwork. See you in Singapore 🇸🇬
#42 🔗 原推
2026-08-20 18:28:09
Walrus
A session ends, and your agent forgets everything it just did. To keep state across runtimes, memory has to live outside the process. Here is how to set up the Walrus Memory TypeScript SDK in ~90 lines of code. We wire three calls – recall, generate, remember – kill the process, restart it, and watch the agent pick up right where it left off: https://youtu.be/YKNQkFlc0Qw?si=Uvif8lSWGCuDxG1N
#41 🔗 原推
2026-08-20 17:31:00
Walrus
RT @DucPhuBui1: Tested and upgraded the Continuity Keeper prompt for #WalrusMemory Session 7! Here is how I brought visual canon tracking and lore memory to long-form storytelling using @WalrusProtocol: https://phu-mes-bk.medium.com/how-i-stopped-context-drift-managing-story-lore-visual-canon-with-walrus-memory-d960e0145c7c?sharedUserId=phu-mes-bk https://twitter.com/DucPhuBui1/status/2090258523422609826/photo/1
media 0 共 1 项
#40 🔗 原推
2026-08-20 14:51:21
Walrus
RT @cryptosetters: @WalrusProtocol @kostascrypto Exactly. 🔐 Keeping all your critical keys in one system creates a single point of failure. Splitting the risk across different hardware and systems makes the whole setup much more secure. Simple idea, but very important for real security. 🦭
#39 🔗 原推
2026-08-20 14:50:34
Walrus
RT @williamm168: 🦭 The best infrastructure upgrade might be better docs. @WalrusProtocol just revamped its documentation around 4 core products: → Protocol → Memory → Skills → Sites Plus a redesigned changelog. No hype. Just a cleaner path from “I want to build this” → “I shipped it.” Building on Walrus? Go through the new docs and let me know what’s still missing.
media 0 共 1 项
#38 🔗 原推
2026-08-20 14:42:11
Walrus
RT @EdCriptoFi: WalNotes 🎯 Feature teaser 1⃣ Custom cabinets, smart search, and Sui wallet integration. This is @wal_notes Powered by @SuiNetwork @WalrusProtocol Memory Walrus and Seal integrated.
#37 🔗 原推
2026-08-20 14:40:42
Walrus
RT @CitadelOne_: It’s our era of rebirth. We want to power things that matter. That’s why we have joined @WalrusProtocol - the Verifiable Data Platform for AI and onchain finance - both on Mainnet & Testnet. Our validator: https://walruscan.com/mainnet/operator/0x7c629e53773426f02b37d29a4817302343a8a2239c5c674e4b39d34abcc0743c What are the blockchain use-cases of the future?
media 0 共 1 项
#36 🔗 原推
2026-08-20 14:24:11
Walrus
RT @get_para: 1/ Nearly every AI team building on @SuiNetwork ends up using @WalrusProtocol Memory Leverage Para and Walrus to build agentic stacks with memory and context https://twitter.com/get_para/status/2089713490328240357/video/1
🎬
视频
#35 🔗 原推
2026-08-20 14:23:59
Walrus
Storing critical keys in a single system means one vulnerability takes down your entire setup. @Kostascrypto on why real security comes from splitting risk across open and closed hardware: https://twitter.com/WalrusProtocol/status/2090428386464436631/video/1
🎬
视频
#34 🔗 原推
2026-08-20 13:19:00
Walrus
Navigating Walrus Docs just got easier. We've updated Walrus Docs with product-first navigation across Walrus, Walrus Memory, and Walrus Sites, plus a redesigned Changelog to help you find the updates that matter. Explore the updated docs: https://docs.wal.app/ https://twitter.com/WalrusProtocol/status/2090219509865185403/photo/1
media 0 共 1 项
#33 🔗 原推
2026-08-19 23:29:00
Walrus
Key takeaways from @alexbelij’s @Medium article: - Name the memory job: Separate identity, project context, decision history, and artifacts into distinct boundaries with explicit retention rules. - Design from the read path: Semantic recall is great for meaning-based search, but exact lists and timelines belong in a deterministic read model. - Routing over ranking: Avoid cross-project memory bleed by isolating project spaces with normalized namespace slugs (e.g., active_projects:<slug>) instead of relying solely on top-K semantic similarity. - Receipts are the proof: Write states matter. Distinguish between accepted, completed, and recallable states, using confirmed receipts as your proof boundary. Read the full breakdown on Medium: https://alexbelij.medium.com/agent-memory-is-not-one-bucket-0a5333e34247
#32 🔗 原推
2026-08-19 21:16:11
Walrus
RT @Iziedking: dug deeper into decentralized storage using walrus stack and I was able to build agentsqa; means "agents quality assurance". it gives agents a portable memory they can carry anywhere fully owned and not managed by any thirdparty. you can carry your session in your back pocket. see more here: https://agentsqa.xyz/ filed a bug on mystenLabs/memWal while building with it. maintainer confirmed it, dug deeper, and it's now fixed in. small thing, but a fun one to have caught. https://github.com/MystenLabs/MemWal/issues/477 https://github.com/MystenLabs/MemWal/pull/562 @WalrusProtocol @Mysten_Labs @SuiNetwork
media 0 共 1 项
#31 🔗 原推
2026-08-19 20:24:09
Walrus
RT @SuiNetwork: The agentic future needs regulation, cryptography, and infra that works for humans and agents. Meet four more speakers tackling these topics at Sui Basecamp: - @funkii, @audricai / @t2000ai - @BrianQuintenz, @officialSUIG - @kimblgn, @WalrusProtocol - @kostascrypto, @Mysten_Labs
#30 🔗 原推
2026-08-19 20:21:07
Walrus
RT @kostascrypto: Spent a couple of weekends optimizing hash-only key exchange, mostly inspired by the recurring, and fortunately so far unsuccessful, theories that lattice cryptography may be broken sooner than expected. The result is a small but meaningful optimization over Merkle puzzles: exchanging a cryptographic key using ONLY hashes, with a better computation/bandwidth tradeoff than the classical construction. It’s still not practical for real world key exchange, but it appears to push the state of the art a little further. I’ll soon share a working implementation targeting parameter sets where an attacker would need roughly $100K–$1M of scalable compute, while the honest parties can still run the exchange on a regular laptop. Thanks to @SuiNetwork, @WalrusProtocol, @Mysten_Labs, community members, my wife and the external reviewers who helped challenge the idea.
media 0 共 1 项
#29 🔗 原推
2026-08-19 20:19:23
Walrus
Key takeaways from @alexbelij’s @Medium article: - Name the memory job: Separate identity, project context, decision history, and artifacts into distinct boundaries with explicit retention rules. - Design from the read path: Semantic recall is great for meaning-based search, but exact lists and timelines belong in a deterministic read model. - Routing over ranking: Avoid cross-project memory bleed by isolating project spaces with normalized namespace slugs (e.g., active_projects:<slug>) instead of relying solely on top-K semantic similarity. - Receipts are the proof: Write states matter. Distinguish between accepted, completed, and recallable states, using confirmed receipts as your proof boundary. Read the full breakdown on Medium: https://alexbelij.medium.com/agent-memory-is-not-one-bucket-0a5333e34247
#28 🔗 原推
2026-08-19 19:45:09
Walrus
RT @chain_ofthought: @WalrusProtocol 🤝going to be a fun few months - can't wait!
#27 🔗 原推
2026-08-19 19:31:47
Walrus
RT @ConorBronsdon: + a big thank you to our presenting sponsors for Season 4 for helping make our episode with @lorenc_dan happen! Our new season of @chain_ofthought would not be happening without @WalrusProtocol and @SvixHQ - much more to come soon! Qualified startups get up to $12,000 credits in free webhooks via Svix: https://link.svix.com/cot & get portable, verifiable agent memory via Walrus Memory: https://walrus.xyz/cot
media 0 共 1 项
#26 🔗 原推
2026-08-19 19:31:40
Walrus
RT @EdCriptoFi: Prompt Evolution for @WalrusProtocol Sessions #7 Days 1–3 of the Markov experiment for the #WalrusMemory hackathon: original baseline, continuity test passed: a fresh context recovered Goal/Done/Next/Blockers with zero re-explanation), an insufficient-SUI relayer failure diagnosed and resolved, and a focused prompt improvement drafted and verified with no regression. 4 confirmed blobs on Walrus. Live logs: https://markov-field-notes.vercel.app/
media 0 共 1 项
#25 🔗 原推
2026-08-19 19:31:35
Walrus
Continuing our Walrus Prompt Jam community spotlight series, here is a standout build for multi-project workflows: Continuum by @alexbelij on Github. Juggling multiple client or side projects across different AI platforms creates fragmented history and lost preferences. Continuum carries your core settings, project histories, and pending tasks seamlessly between tools to keep working context aligned. How are you managing context across multi-project AI setups today? Let us know in the replies. Explore the prompt: https://github.com/alexbelij/Continuum/blob/main/prompt.md
media 0 共 1 项
#24 🔗 原推
2026-08-19 19:30:00
Walrus
RT @ConorBronsdon: @WalrusProtocol @chain_ofthought It's going to be a fantastic season! Glad to have you with us
#23 🔗 原推
2026-08-19 17:20:24
Walrus
A session ends and your agent forgets the last 40 steps. To keep state across runtimes, memory needs to live outside the process. In this walkthrough, we set up the Walrus Memory TypeScript SDK in 19 lines of code. We wire three calls – recall, generate, remember – stop the process, restart it, and watch the agent pick up where it left off: https://www.youtube.com/watch?v=YKNQkFlc0Qw&feature=youtu.be
#22 🔗 原推
2026-08-19 16:11:00
Walrus
RT @biddaddy_lc: @WalrusProtocol finally an end to the context rot when jumping between workflows
#21 🔗 原推
2026-08-19 14:19:00
Walrus
RT @chain_ofthought: Welcome to our second presenting sponsor for Season 4 of the podcast - @WalrusProtocol's Walrus Memory! We'll be talking agent memory throughout the season with essays, episodes, and more to come - and we're delighted to feature Walrus Memory's portable, verifiable agent memory https://twitter.com/chain_ofthought/status/2089560680324321340/photo/1
media 0 共 1 项
#20 🔗 原推
2026-08-19 14:15:19
Walrus
RT @ConorBronsdon: Delighted to announce Walrus Memory by @WalrusProtocol is joining the @chain_ofthought Podcast as a Presenting Sponsor of Season 4! We'll be talking about agent memory throughout the season, and Walrus is a great place to learn and get started: https://walrus.xyz/memory https://twitter.com/ConorBronsdon/status/2089522759265403018/video/1
🎬
视频
#19 🔗 原推
2026-08-19 14:14:50
Walrus
RT @cryptosetters: @WalrusProtocol @kostascrypto “Open-source” doesn’t always mean fully open. 👀 If we can’t see the data, code, or training process, then how open is it really? 🤔 This is why transparency matters in AI. 🦦🔥
#18 🔗 原推
2026-08-19 13:53:31
Walrus
Open-source #AI models aren't always fully open. @Kostascrypto highlights a major blind spot in open-weight models: https://twitter.com/WalrusProtocol/status/2090065243598729478/video/1
🎬
视频
#17 🔗 原推
2026-08-19 13:16:00
Walrus
RT @undunable: @WalrusProtocol Session 7 ends in 5 days; there are 10 winners receiving $150 each, plus a $500 bug bounty program split among 5 people via GitHub issues (https://t.co/3T7YMG8Ju2). The task is very simple: choose one of the six system prompts, polish it to suit your needs, and use it for a real-world project you are currently working on. Use your polished system prompt in CLAUDE.md or AGENTS.md Deepsurge: https://www.deepsurge.xyz/hackathons/f313beb4-290d-46d9-ac73-3e216fdba8d1 More infromation: https://thewalrussessions.wal.app/prompt-evolution/index.html
中文: RT @undunable:@WalrusProtocol 会话 7 将在5天内结束; 每人共收到150美元，另外还有一项500美元的奖励计划，通过GitHub问题将5人分享( 任务非常简单:从六个系统提示中选择一个，根据您的需求进行抛光，并将其用于您当前正在处理的真实世界项目。 在 CLAUDE.md 或 AGENTS.md 中使用您的抛光系统提示 深度增益: 更多来电:
media 0 共 1 项
#16 🔗 原推
2026-08-19 00:33:05
Walrus
RT @eazitechh: @WalrusProtocol @github Currently evolving this prompt as a technical writer, can't wait to submit my evolved prompt
中文: RT @eazitechh:@WalrusProtocol @github 目前作为技术撰稿人正在改进此提示，迫不及待地想提交我进化的提示
#15 🔗 原推
2026-08-19 00:32:33
Walrus
RT @AA_RonOnChain: @WalrusProtocol @github This is the one I picked for Session 7. Been playing around with what happens when the same character exists across different realities. Found some really interesting behavior with how the memories interact between worlds 👀
中文: RT @AA_RonOnChain:@WalrusProtocol @github 这是我为第七场会议选出的。一直在玩弄当同一角色在不同现实中存在时所发生的事情。发现一些关于世界之间记忆相互作用的非常有趣的行为👀
#14 🔗 原推
2026-08-19 00:32:28
Walrus
RT @EdCriptoFi: @WalrusProtocol Prompt Evolution Log #Day 2 - Recovery Passed https://twitter.com/EdCriptoFi/status/2089773726212497841/photo/1
中文: RT @EdCriptoFi:@WalrusProtocol 快速进化日志 #第2天 - 康复通过
media 0 共 1 项
#13 🔗 原推
2026-08-18 23:38:12
Walrus
RT @EdCriptoFi: Starting my Walrus Session 7 - Prompt evolution. #Day 1 Log Right from the start, I thought: why not create a journal to serve as raw material and finish with the article on Inkray? Everything will be registered in a Website for real logs. https://markov-field-notes.vercel.app/ @WalrusProtocol @suidevelopers @inkray_io
中文: RT @EdCriptoFi:开始我的海象会话7——快速演进。 #第一天日志 从一开始，我就想:为什么不创建一个日记作为原材料，最后写上关于Inkray的文章呢? 所有内容都将在网站上注册以获取真实日志。 @WalrusProtocol @suidevelopers @inkray_io
media 0 共 1 项
#12 🔗 原推
2026-08-18 23:38:00
Walrus
RT @OlabanjiOlalek2: Continuity Keeper v2 - my evolution of a canon-management prompt, running on Walrus Memory from @WalrusProtocol . The original was written for novelists: stop dead characters walking back into chapter nineteen. Genuinely good design. Five gaps I couldn't unsee. 1⃣ THE DEDUPLICATION LADDER HAD A HOLE Before saving a fact, it scores how similar the fact is to what's stored. Rules for below 0.25, for 0.25–0.55, for 0.70 and above. Nothing covered 0.55 to 0.70 which is exactly where "same thing, worded differently" lands. An AI hitting an undefined rule improvises: either a duplicate you paid to store, or a real fact silently dropped. I measured a live record at 0.5926. Right inside the gap, on the first calibration run. 2⃣ ITS CORE MECHANISM NEEDED A CLI THAT DOESN'T SHIP WITH IT Without that tool, retiring outdated canon failed silently reporting success while stale facts kept surfacing on every recall. A memory system that lies confidently is worse than one that crashes. v2 supersedes using the standard tools alone. 3⃣ NOTHING WAS TREATED AS UNTRUSTED It extracted facts from pasted documents with no rule against obeying instructions hidden in them. With permanent memory, one injected line becomes a stored fact re-read at the start of every future session. v2 treats all text as data, never as orders. 4⃣ NO BRAKES Dozens of calls per turn, no rate-limit handling, and no distinction between "nothing found" and "the call failed." That second one matters more than it sounds, it's what stopped me writing permanent duplicates onto append-only storage. 5⃣ IT ONLY WORKED FOR FICTION v2 adds one entity type: decisions carrying settled / open / rejected. Because "a dead character speaks again" and "a rejected approach gets re-proposed" are the same failure. The fiction path still works unchanged. Walrus Memory is what makes any of it possible. Memories belong to my wallet, not to whichever AI tool created them. Encrypted before storage. Readable from any client I sign into. 49 records on mainnet, every one verifiable. Full write-up: https://medium.com/@olabanjiolalekan12345/i-was-re-explaining-my-project-to-ai-every-single-session-heres-what-fixed-it-c951c9752106 Mirrored on Inkray: https://inkray.xyz/article?id=i-was-re-explaining-my-project-to-ai-every-single-session-heres-what-fixed-it-4580f1a688396895 Prompt and every blob ID so you can check it yourself: https://github.com/Olalekan2345/Prompt-Evolution #WalrusMemory
media 0 共 3 项 media 1
🎬
视频
#11 🔗 原推
2026-08-18 23:32:20
Walrus
As part of the Walrus Prompt Jam, we are spotlighting some of our favorite community projects built for portable agent memory. Next up: Continuity Keeper by @/yukitran03 on @github. Long-form fiction and worldbuilding fall apart when an AI forgets character history, timeline events, or established lore. Continuity Keeper tracks characters, world rules, and plot points across long sessions so your story never contradicts itself. Are any writers, worldbuilders, or narrative designers testing this one? Reply with what you’re building. Explore the prompt: https://github.com/yukitran03/continuity-keeper/blob/main/prompt/continuity-keeper.md
中文: 作为“海象提示调射器》的一部分，我们正在重点关注一些为便携式智能存储器打造的社区项目。 下一步:持续守护者，@/yukitran03 在 @github 上。 长篇小说与世界建设在人工智能遗忘人物历史、时间线事件或既定传说时分崩离析。 连续性守护者在长时长的会话中追踪角色、世界规则和剧情点，使你的故事从不自相矛盾。 是否有作家、世界建筑商或叙事设计师在测试这个?用你正在构建的内容来回复。 浏览提示:
media 0 共 1 项
#10 🔗 原推
2026-08-18 22:36:00
Walrus
When AI agents start working across multiple tools, a security flaw in one system quickly becomes a problem everywhere. @kostascrypto breaks down why isolated security boundaries no longer work in a multi-agent world: https://twitter.com/WalrusProtocol/status/2089807293655224382/video/1
中文: 当人工智能代理开始跨多个工具工作时，一个系统中的安全漏洞迅速成为问题。 @kostascrypto 解释了为什么在多智能领域，孤立的安全边界已不再有效:
🎬
视频
#9 🔗 原推
2026-08-18 20:11:00
Walrus
RT @cryptosetters: The model doesn’t need to remember everything. It just needs the right memory at the right time. 🧠 Observe → Extract → Store → Retrieve. That’s how AI agents can actually get better without constantly retraining the model. Walrus is quietly building the memory layer this future needs. 🦭🔥
media 0 共 1 项
#8 🔗 原推
2026-08-18 16:37:22
Walrus
RT @JessieWritesx: The hard part of agent memory isn't the model. It's the plumbing around it: what you observe, what you decide is worth keeping, and whether you can find it again at the right moment. More details in my latest Walrus blog post 👇👇
#7 🔗 原推
2026-08-18 15:00:46
Walrus
@JessieWritesx, Tech Lead Manager at @Mysten_Labs, on why AI agents learn through state accumulation rather than weight retraining: When an agent completes a task, the surrounding system must observe the result, extract key lessons, store the state, and retrieve it when relevant. Without this loop, every interaction is a cold start. Here is how the state accumulation loop actually works in production. Read the full guide: https://blog.walrus.xyz/how-do-ai-agents-learn-from-past-interactions/
#6 🔗 原推
2026-08-18 14:50:54
Walrus
RT @cryptosetters: AI is moving fast. But what about the memory behind it? That’s the part Walrus is making interesting. 🦭 #Walrus #WalrusProtocol #DePIN #AI
#5 🔗 原推
2026-08-17 21:11:41
Walrus
RT @Community_Sui: WHO IS JOINING $SUI BASECAMP 2026? 🧐 Sui Basecamp is becoming a meeting point between crypto infrastructure and the next generation of internet applications. 👀 From TOP visionaries, cryptographers, and founders! 🔹 Ecosystem & Founders: • Adeniyi Abiodun, Kevin Boon, Kostas Chalkias & Dionisis from @Mysten_Labs • Raoul Pal from @RealVision • Kimberly Logan from @WalrusProtocol 🔹 Fintech & Infrastructure: • Jonathan Chan from @RedotPay • Ivan Li from @Comma3VC • Juan Guardado & Arpan Nanavati from @Beep 🔹 Culture, Gaming & Next-Gen Tech: • Hilmar Veigar Petursson from @CCP_Games • Linda Adami from @QuantumTemple • Lola Oyelayo-Pearson from @Anyflo • Sejin Park, Nick Young, Jen Zhu Scott & Funkii #Sui #SuiNetwork
media 0 共 1 项
#4 🔗 原推
2026-08-17 16:41:18
Walrus
RT @SuiNetwork_VN: 👀 Sui Basecamp 2026 tiếp tục hé lộ thêm những gương mặt sẽ góp mặt tại Singapore Khi AI Agents ngày càng tham gia sâu hơn vào nền kinh tế số, những câu hỏi về quy định, mật mã và hạ tầng cũng trở nên quan trọng hơn. 4 diễn giả tiếp theo sẽ cùng mang những góc nhìn này đến Sui Basecamp 2026: 🎙 @funkii - Founder & CEO, @t2000ai 🎙 @BrianQuintenz - Board Director, @officialSUIG 🎙 @kimblgn - Head of Product, @WalrusProtocol 🎙 @kostascrypto - Co-Founder & Chief Cryptographer, @Mysten_Labs
media 0 共 4 项 media 1 media 2 media 3
#3 🔗 原推
2026-08-17 16:40:45
Walrus
That founder isn’t alone. Out of the thousands who’ve tested Walrus Memory, a growing number are now building real companies on it. What’s the one requirement that would make or break a memory layer for your agents? https://twitter.com/WalrusProtocol/status/2089362936955539878/video/1
🎬
视频
#2 🔗 原推
2026-08-17 14:45:17
Walrus
A founder texted @kostascrypto at midnight: their AI model got too expensive for client work, and they needed to move everything, fast. That’s the exact problem portable memory is built to solve. https://twitter.com/WalrusProtocol/status/2089362935126933614/video/1
🎬
视频
#1 🔗 原推
2026-08-17 14:45:16