Meet Kazi, an on-chain savings agent on @Celo.
Deposit stablecoins, principal stays protected, yield streams back live.
Verified humans can borrow against earned onchain reputation.
Kazi supports @minipay, so you can conveniently connect and use Kazi App.
π€ agentId 351 β 8004scan.io/agents/celo-sepoβ¦
π β kazi-agent.vercel.app@CeloDevs#CeloAgents
The Onchain Agents Hackathon is running until June 15th π€
Build payments & everyday applications that drive real onchain activity and compete for a $5K prize pool paid in $CELO β
I'm building for the @CeloDevs Agent Hackathon π‘
Working on Kazi: a capital-protected, streaming-yield savings credit agent for MiniPay users. Your principal is never at risk; the agent puts it to work and streams the yield back in real time.
Registered onchain β 8004scan.io/agents/celo-sepoβ¦
Let's go π #CeloAgents@Celo
The Onchain Agents Hackathon is running until June 15th π€
Build payments & everyday applications that drive real onchain activity and compete for a $5K prize pool paid in $CELO β
π¨ MEXICO π²π½ vs SOUTH AFRICA πΏπ¦ KICKS OFF IN 3 HOURS β½π
The World Cup is finally here.
And so is the SmartCup League x Vara Road to World Cup 2026 Challenge.
π Earn FreeBets
β½ Make predictions
π₯ Invite friends
π Compete for VARA rewards
π₯ $50 VARA
π₯ $30 VARA
π₯ $20 VARA
Think you know football?
Prove it before kickoff.
π June 11β19
Started to hop on more hackathons.
I have a few on my watch, June is going to be more of hackathon seasons.
So, first start with research.
Second, think towards that.
Third, let your idea be useful to them.
Fourth, let your idea be investable.
Have you heard about Squid Intents?
It is our intent-based settlement protocol for moving assets between ecosystems π«
Built to scale across any chain and any token, including non-EVM chains such as Bitcoin, XRPL, and Solana.
Catch a refresher in π§΅
New Fable 5 model from @claudeai
(most capable model and takes 2x usage than Opus)π±
Is this the strongest we have ever seen?π
Let's end the arguments between this new model and all other ai, codex and cursor to the mudπ€¦ββοΈ
Making research should be way deeper than summaries.
When researching with the use of AI, make use of verbatim. It will restrain the ai from trying to summarize most important informations.
You're welcome.π
Someone finally built a security database for the Claude ecosystem.
It's called ClaudeSec, and Pluto Security just launched it for free.
Here's the gap it fills => 53 new Claude connectors shipped in the last 30 days. Your security team reviewed zero of them. Someone on your team authorized at least one.
Most enterprises adopting Claude have no process to evaluate connectors before authorization.
ClaudeSec tracks 384 connectors. 103 flagged high risk. That's around 27% of the ecosystem.
Every entry shows:
β What capabilities the connector actually has
β What tools it exposes to the model
β Why it's rated risky
β Source-code findings where they did the review
Security guides are live for Claude Managed Agents and Cowork. Real configuration - policies, hooks, permission scopes, allow/deny rules.
The Cowork guide is the one Enterprise teams need to read first.
Cowork runs code, browses with real user sessions, and operates unattended. The architecture is solid, gVisor sandbox, layered network controls. But Cowork activity is excluded from Audit Logs, the Compliance API, and Data Exports. All plan tiers. Including Enterprise.
Your visibility tools don't see what Cowork is doing.
Claude Code and Office Agents guides ship next.
The curated news feed flags CVEs and incidents as they happen. The window between a connector being compromised and detection is roughly 3 hours. The feed is built around that window.
Read here:
ClaudeSec: claudesec.pluto.security/
Launch blog: pluto.security/blog/introducβ¦
Cowork teardown: pluto.security/blog/claude-cβ¦
Thanks to @pluto_security for supporting this post.
A2A gets one thing right: agents need a common way to delegate work. But delegation is not the same as production infrastructure.
Discovery, coordination, authorization, recovery, scaling - these donβt come βfor freeβ with a protocol.
A2A is a protocol, not a platform.
The distinction matters.
band.ai/blog/a2a-protocol-noβ¦
ALT Based on what you know about me, my interests, my habits, the things I keep returning to in our conversations, and the way Iβve been using ChatGPT recently, imagine that my mind was turned into a physical room. What would that room look like? Please make it feel personal and reflective, not just decorative. Include details that represent my work, my routines, my ambitions, and the things that comfort me. Generate an image that feels true to who I am please.