Dappnode retweeted
MetaMask Staking doesn't control the withdrawal keys. an operator compromise can't redirect the staked ETH.
it can redirect the block fees.
Sep 30, 12:12–16:46 UTC: 18 blocks from @MetaMask -run validators paid their fees to a wallet funded through Tornado Cash at 10:27.
total: 0.36 ETH.
whoever set that address had access somewhere inside the validator operation.
that's enough to treat the signing environment as potentially compromised. bad signatures can get validators slashed.
Ethereum can't rotate a validator signing key in place. you exit, withdraw and restake with fresh keys.
so they're rotating the validator.
btw, we've seen the containment playbook before.
Sep 9, 2025: Kiln disclosed a security incident and exited its Lido validators.
28 days later, the cause: compromised GitHub token -> CI workflows -> cloud credentials.
104 days later, the bill: 207.312 ETH in missed rewards across 5,726 validators.
MetaMask hasn't disclosed the root cause or the validator count.
one thing is already different. Kiln found no evidence its Ethereum validators were touched.
here, someone was already collecting the fees.
etherscan.io/address/0x98B92…
Security Update: We are responding to a security incident affecting part of our infrastructure.
At this time, we have identified no immediate threat to MetaMask wallets.
As a precaution, we are proactively exiting affected validators within our non-custodial staking operations, in coordination with clients, partners and security advisors.
We’ll share further updates as appropriate. metamask.io/news/user-update…
If you are not running your own agent you are behind
"But it's too complicated"
In 1:47 seconds you can have it running on your Dappnode
Get your agent, fully private & self-sovereign
Comment 👇which thing you'd die to automate in your life and we'll make a video of how
Dappnode retweeted
I almost missed the IMHO most catastrophic @ledger vulnerability:
h/t @Fatalmeh who brought it onto my feed
Dappnode retweeted
kinda crazy my @dappnode video from 2023 is still one of my top performers on youtube
youtu.be/VR_aZXtX3r4?si=2wlT…
Dappnode retweeted
The guys from @exolabs , which we have used extensively in @dappnode, dropped local.ai.
For virality, they let you claim your name already.
local.ai/lanski/invite
Testing on Dappnodes
Today, we're introducing Mach-1 Additive, a 35 billion parameter model that can inference without ever multiplying by a weight. At 1.7 bits per weight, Mach-1 recovers 95% of the performance of the original full precision model, Qwen 3.6 35b, across 12 agentic and reasoning benchmarks, while being 10x smaller.
At 7GB, Mach-1 comfortably fits on consumer laptops with speeds of up to 120 tokens per second, making local inference not just feasible but useful.
Unlike algorithms like BitNet, our approach requires minimal retraining, under 15 GPU hours, making it scalable to massive LLMs. Over the coming weeks, we will be announcing and serving models of up to 3 trillion parameters compressed using our algorithm. For now, you can visit our website to play with Mach-1 directly in your browser, or download our desktop app.
We couldn't be more excited to launch Mach-1. We're looking forward to an energy efficient future for AI, powered by scaled intelligence density.
Dappnode retweeted
2. CSMv3: Expanding Permissionless Staking
Through this upgrade Lido CSM becomes even more reliable and operator-friendly.
Alongside multiple under-the-hood technical improvements, the primary new features are:
• New Node Operator type: Identified DVT Clusters
• Native Node Operator reward splitting
• Permissionless stake share increase through the lightweight Easy Track governance














