Nostr notes by aljaz

and i thought i'm bad at naming things

2026-10-27T18:16:40Z

In reply to nevent1q…53m5
_________________________

and i thought i'm bad at naming things

Attested #2 is out - weekly newsletter about confidential ...

2026-10-27T15:42:41Z

Attested #2 is out - weekly newsletter about confidential inference and confidential computing in general

quoting
naddr1qv…r9d2

Phala’s guest-agent incident: the enclave held, the plumbing didn’t

Phala published a post-mortem on Oct 19 for an attack on Phala Cloud between Oct 8 and Oct 12.

Short version. dstack runs a guest-agent inside every CVM, and Phala’s gateway exposes its read-only endpoints publicly so anyone can fetch the TDX quote and measurements. Same agent also served container logs. In dstack OS images v0.3.1 through v0.6.0 the /logs/<container> endpoint passed the container name straight into the Docker API path, so a crafted name turned a log read into a container inspect, which returns Config.Env. Environment variables. Where a lot of apps keep their secrets.

The attacker started from inside a customer app that runs user-submitted code, reached an unauthenticated host management API over the CVM’s user-mode network, changed firewall rules on that host, pulled a shared gateway admin token out of the gateway container’s config and used it to make 63 changes to DNS-credential and ACME settings on one gateway. Separately, it read logs and env vars from other tenants’ apps over the public route. Phala caught it on Oct 12 at 19:28 UTC, about 21 hours after the first scan from inside that app (Oct 11, 22:03, per its timeline), when a certificate reconciliation check started failing on the tampered gateway.

What Phala says was not touched: TEE and KMS keys, sealing keys, the on-chain KMS, customer disks, host SSH. Fixes shipped Oct 13 in dstack v0.6.1 and v0.5.12. Yes, the same v0.6.1 I told you to patch to in #1 for the SEV-SNP certificate bug. Turns out it fixed more than that. Everyone has to be on a fixed image by November 19; from Nov 20 Phala auto-upgrades remaining v0.5.x/v0.6.0 CVMs and retires v0.3.x ones (enterprise plans are exempt from that retirement). Notified customers should rotate whatever was in their env vars.

Why it matters

Nothing here broke TDX. It broke the management agent, the host network isolation and a token that several gateways shared, plus a console template that set public_logs=true by default while dstack itself defaults it off (Phala lists making it opt-in as still in progress). And the vulnerable guest-agent was part of the measured image: genuine, attested code doing exactly what it was built to do, badly. The secrets were sitting in the env of the workloads, readable through a side door. If your threat model ends at “it’s in a TEE”, this is the counterexample. Credit to Phala for publishing this much detail: timeline, root causes, the lot.

If you run on dstack yourself: check your OS image version and whether public_logs is on.

DeepSeek 1017 spreads. Read the fine print

In #1 only Tinfoil had a measured 1017 build. A week later most providers I track list it, and the labels mean different things (model APIs checked Oct 27):

Tinfoil underneath: Redpill and NanoGPT both list theirs as running through Tinfoil (Redpill model API). Phala says it’s “served inside a TDX-attested GPU TEE”. I checked on Oct 27: the route I tested is Phala’s TDX gateway passing requests (signed receipt and all) to Tinfoil, whose GPU enclaves matched Tinfoil’s signed 1017 build, so attested yes, a Phala GPU TEE no.

Not attested: NEAR AI serves it through OpenRouter ZDR (verifiable: false in its model API), Chutes lists it as non-TEE pay-as-you-go (model API), and Venice and PPQ’s explicit 1017 ID aren’t TEE routes either.

Maple and Prem list it as confidential. My monitor has no per-model check for Maple yet, and Prem’s 1017 route passes hardware and TCB but isn’t tied to a measured 1017 build (Oct 27). Correction to #1: Prem’s free cached tokens on confidential models now run to Dec 1, per its model table.

Why it matters: “available on X” isn’t “available inside X’s enclave”. The model API usually tells you which one you’re getting.

What we measured

Catalog changes this week

Chutes is ending monthly plans and adding pay-as-you-go models that aren’t TEE (Oct 26 post), and says TEE inference continues. Four of its TEE models were also gone from its public catalog in my Oct 27 check, unexplained so far.

Tinfoil and NEAR AI added “decision” models that return probabilities and scores instead of text (Tinfoil model catalog).

NEAR AI‘s new GLM 5.3, listed as attested, fails my check: the challenge nonce doesn’t come back, so I can’t tell a fresh quote from a replayed one (NEAR attestation page, Oct 27).

Post-quantum TLS

In the Oct 27 checks, 10 of 11 provider endpoints negotiated hybrid X25519MLKEM768.

New since last week: Chutes, now does on llm.chutes.ai. NEAR AI is the remaining one with no hybrid group.

All 11 still accept a classical handshake, and for all ten hybrid ones the site marks the result edge-only: I can’t prove TLS terminates inside the attested TEE. So still nobody, on current data, with a post-quantum TLS channel that provably ends in the enclave.

Elsewhere

Cloudflare leaked prompts between customers. Its post-mortem (Oct 26): between Aug 27 and Sep 9, 155 responses from GLM 5.3 on Workers AI contained pieces of other customers’ recent requests, all to prompts over 681,280 tokens. The prefill GPUs’ KV cache pool was smaller than the advertised 1M context, and in split prefill/decode mode SGLang passed a rejected request on to decode anyway, which read pages still holding earlier requests’ data. Cloudflare says it found no evidence of deliberate exploitation; the upstream fixes shipped in SGLang v0.5.20 through v0.5.22. Why it’s here: an enclave isolates the host from the workload, not one request from the next. If many tenants share one attested engine, a bug like this leaks just the same, inside a perfectly valid attestation. Same lesson as Phala, one layer up.

Tinfoil threw out OVMF. Their writeup (Oct 25) replaces about a million lines of C firmware with roughly 300 lines of assembly. Page tables, ACPI tables, kernel, initramfs and command line are built ahead of time and measured at launch (MRTD on TDX, launch digest on SEV-SNP), so the firmware no longer measures anything and only checks two host values, CPU count and memory map. Their argument: build tool bugs are in the audit TCB, runtime parser bugs are in the attack surface. I like it. It’s also an honest trade, their firmware refuses to boot machine shapes they didn’t build for. The code is public on a branch of cvmimage; the post doesn’t say whether it’s what production runs today. Timely too, given the EDK II pre-boot RCE in the PXE driver published Oct 22. Why does VM firmware have a network stack? Exactly.

Google made Confidential G4 GA. NVIDIA RTX PRO 6000 Blackwell with NVIDIA confidential computing, on Compute Engine and GKE, in 36 zones (Google Cloud blog, Oct 21). Confidential Space support for it is “rolling out to general availability in the coming weeks”. More rentable confidential GPU capacity is what this market is short of.

Cohere put Model Vault behind attestation. Confidential Computing on Model Vault is in public beta since Oct 22: H100s in confidential mode on Azure, requests encrypted end to end into the enclave, attestation the customer checks before sending (Cohere blog). Free for design partners during the beta, otherwise “contact Cohere” for rates, which keeps it off my price comparison. Same week, NVIDIA announced Confidential Bare Metal (Oct 21), for model builders who want to run their own host stack on third-party GPUs. Cohere says it plans to use it, and Google says it’s bringing it to its GB300 instances.

Microsoft said the quiet part. Its Azure + NVIDIA + Protopia reference architecture (Oct 19) has this line: “an attested GPU does not establish how a gateway or cache handles requests.” See the two items above.

Missed last time: OPAQUE announced on Oct 13 that its team is joining Anthropic, and its site now says OPAQUE Systems is winding down; its open-source projects continue under AgenTrust. Reporting (not the companies’ own posts) says Anthropic licensed the technology; neither has said if or how it ends up in Claude.

Releases. Privatemode v1.60.0 (Oct 22) fixes GLM-5.3 empty responses and moves to Contrast v1.25.1, a patch release that fixes govulncheck findings and blocklists the compromised eMudhra certs. NanoGPT added Tinfoil’s PII redaction (Oct 27), with detection running in an attested enclave before your text reaches the model provider. tinfoil-go shipped v0.16.4 and v0.17.0 release candidates. Nothing new from nvtrust.

New names I’m looking at

Surplus (surplusllm.com) says private inference is live (Oct 21): 22 confidential models on Intel TDX + NVIDIA confidential GPUs through Phala, prompts encrypted client-side (if you use its E2EE headers; without them, its own privacy page says the gateway relays plaintext). Per its X bio it’s paid in USDG on Robinhood Chain, so it’s another crypto-paid router on top of Phala, same pattern as Umbra in #1. Not verified.

VoltageGPU (voltagegpu.com) is mostly a different kind of thing. Its site says it rents Intel TDX confidential VMs with H100, H200 and RTX PRO 6000 GPUs, that tenants generate the TDX quote themselves on its Confidential VM tier, and that it has run NVIDIA GPU attestation from inside its VMs on some of those configurations (not yet on B200/B300). That’s the “run it yourself” option. It also advertises an OpenAI-compatible API with 10 TEE models, which I’ll look at separately before it goes anywhere near the price comparison. Not verified either.

$1 million fund for products running on .tech domains ...

2026-10-27T06:40:27Z

$1 million fund for products running on .tech domains
https://noexcuses.get.tech/

EVOO + free range eggs = no seed oil mayo

2026-10-26T07:03:51Z

In reply to nevent1q…9yp5
_________________________

EVOO + free range eggs = no seed oil mayo

New unlock for old cards to give them a lot more performance ...

2026-10-26T06:58:06Z

New unlock for old cards to give them a lot more performance https://github.com/Zagitta/gpu-unlock

if anyone needs a project away from the keyboard this looks fun.. ...

2026-10-22T07:38:23Z

if anyone needs a project away from the keyboard this looks fun..

the state of cybersecurity ...

2026-10-22T05:14:13Z

the state of cybersecurity

its really rewarding to maintain a lot of websites and ...

2026-10-21T17:42:59Z

its really rewarding to maintain a lot of websites and infrastructure in 2026

it is one of the providers on confidentialinference.net 😉

2026-10-21T17:33:28Z

In reply to nevent1q…rhrv
_________________________

it is one of the providers on confidentialinference.net 😉

i started a weekly newsletter to force myself to keep up with ...

2026-10-21T14:48:56Z

i started a weekly newsletter to force myself to keep up with whats happening in the confidential computing space and to keep https://confidentialinference.net/ updated

quoting
naddr1qv…qq3a

So here’s the first issue of Attested. I’ve been running confidentialinference.net for a while now, mostly because friends kept asking me which confidential inference providers to use (or didn’t know there were any). A weekly writeup felt like the obvious next step. Or a very elaborate way to procrastinate, I’m not sure which one is actually true. But it also gives me an excuse to keep myself up to date on all things confidential inference (and computing).

Brave Leo now talks to a NEAR enclave

On Oct 15 Brave turned on end-to-end encrypted chat for GLM 5.3 Flash in Leo (Release, Beta and Nightly). Your prompt gets encrypted with Oblivious HTTP to a key that lives inside a NEAR AI TEE running TDX plus NVIDIA confidential computing. Brave’s relay sees your IP but not the prompt. NEAR’s enclave sees the prompt but not your IP.

Before anyone gets too excited - the prompt still gets decrypted. The model has to read it. What you get is operator isolation: the hardware is supposed to keep infrastructure operators (NEAR and Brave included) from reading what the attested code inside the enclave is processing.

The browser only gets the enclave key after Brave’s open source verifier checks the TDX quote, GPU attestation and a rebuild of NEAR’s code. Today that check runs in Brave’s gateway, not your browser. Brave says so itself and lists in-browser verification as a next step. And its opt-in, you only get it if you select GLM 5.3 Flash.

Its not the first confidential inference in production, Apple’s Private Cloud Compute was earlier. But a mainstream browser asking for an attestation quote instead of a pinky promise is the direction I want.

On my NEAR attestation page GLM 5.3 Flash is the only one of NEAR’s 9 TEE models that gets past the hardware check. The other 8 fail it.

For the harvest-now-decrypt-later crowd: OHTTP here is HPKE with X25519, which is classical, and NEAR’s public API endpoint didn’t negotiate hybrid post-quantum TLS in my latest probe, PQC run #32.

Tinfoil had DeepSeek’s new weights in a measured build the next day

DeepSeek published DeepSeek-V4.1-Flash-1017 on Oct 17 and on Oct 18 Tinfoil switched the B300 build in its enclave repo to the new weights, pinned by commit hash to the same revision Hugging Face serves (1bae0e10…).

Each release ships with SEV-SNP and TDX measurements, the digest of its deployment manifest and a Sigstore entry for that digest. The container image has its own digest, pinned separately in the build’s config. Both v0.0.6+b300 and v0.0.7+b200 are up in the enclave repo.

This is how a model update on an attested service should look. Hash in a public config, a measurement to compare against the attestation report, an entry in a transparency log. No “trust me bro, we updated the model”.

Heres the catch tho. A public release doesn’t prove a given production request ran on that build, only fresh attestation from the serving endpoint matched to the release measurement does. I don’t have that yet. Tinfoil kept the public model name, deepseek-v4-1-flash, still at $0.65/$1.45 per M input/output in its model catalog, and my monitor couldn’t fetch fresh evidence for that route in its latest run (see the Tinfoil attestation page). So no match to a release. Yet.

Prem also lists “DeepSeek V4.1 Flash · 1017” among its confidential models, at $0.35/$1.40. Thats Prem’s own listing and my monitor hasn’t confirmed which checkpoint its enclave serves. Provider listing, not a deployment anyone has verified, unlike Tinfoil’s measured build. Its also the cheapest confidential price for that model I track, vs Tinfoil’s $0.65/$1.45 which PPQ and NanoGPT resell. And the same page has cached input on confidential models free until October 20. Which is the day this goes out, so don’t plan around it.

Chutes quietly cut Kimi K3 by 38%

Chutes cut prices on four models on Oct 12 and didn’t announce it anywhere I could find. Kimi K3 went from $3.00/$15.00 to $1.85/$9.50 per M input/output tokens, so 38% off input and 37% off output. Date comes from archived copies of their own pages: the model API at 05:31 on Oct 12 still had the old prices, the pricing page at 19:24 the same day had the new ones (Rome time).

Pricing in this space is still very .... varied. Kimi K3 elsewhere, public per-token TEE prices per M in/out, checked Oct 19:

NanoGPT: $3.00 / $15.00

Redpill: $3.00 / $15.00

NEAR AI: $3.30 / $16.50

Venice: $3.75 / $18.75

Tinfoil: $4.00 / $20.00

PPQ.AI: $4.22 / $21.10

The Redpill one is my favorite. Its a TEE route that Redpill’s own API says is served by Chutes, at a 62% markup on Chutes’ direct input price. Reselling with a markup, as one does.

What my monitor saw

Prem joined on Oct 17 with the full check. First results from the Oct 19 run: 2 of 5 models pass hardware and TCB, DeepSeek V4.1 Flash and Nemotron 3 Ultra, both on Intel TDX. Both stall at channel binding. My monitor couldn’t verify that the encryption key is bound to the attested environment, so it can’t confirm the attested E2EE Prem documents and files both as “TEE-only gateway”. Thats a gap in what I could establish. Its not evidence that Prem’s gateway sees plaintext. The other 3 have been unreachable since Oct 17, so no verdict, good or bad.

Now the number thats gonna get screenshotted. Of the 97 checks my monitor runs on 11 providers, 55 pass hardware and TCB and 0 pass every gate my policy applies, as of the Oct 19 run.

Before somebody takes that out of context: its about my verifier, the routes I test, my policy and this observation window. It doesn’t prove that no provider can be verified end to end.

Almost everyone stops at workload identity (15 of 97), source provenance (1 of 97) or request route (0 of 96). Privatemode gets 13/13 through hardware and TCB, Chutes 13/14, Tinfoil 9/9 (tho 8 of Tinfoil’s had no fresh evidence in the latest run). NEAR AI gets 1 of 9, the Brave model.

Post-quantum TLS, from PQC run #32 on Oct 18: 9 of 11 provider endpoints negotiated hybrid X25519MLKEM768. Chutes and NEAR AI didn’t, and all 11 still accept a classical handshake. The nine hybrid ones are marked edge-only, because the monitor hasn’t established that TLS on those tested endpoints terminates inside the attested TEE. TLS only tho, application-layer encryption is separate. For Chutes the attestation dashboard records ML-KEM-768 application encryption keys.

Two software-only breaks of SEV-SNP

FABRICKED and STALEUS, both from Benedict Schlüter and Shweta Shinde’s group at ETH Zurich, published at USENIX Security 2026 (the publisher dates it to August 2026). Both hit AMD SEV-SNP without physical access.

FABRICKED needs an attacker controlling both the UEFI firmware and the hypervisor and ends in arbitrary read/write in the guest plus forged attestation reports (CVE-2025-54510).

STALEUS needs hypervisor privileges and gets arbitrary read/write inside a fully attested CVM (CVE-2025-54509).

“The hardware attested” only means something if the TCB version behind it has the fix. On my monitor Privatemode and Tinfoil attest on SEV-SNP, and so do the PPQ and NanoGPT routes that resell Tinfoil. Thats why TCB policy is its own column on the attestation pages.

Patch your dstack

Phala’s dstack v0.6.1 (Oct 13) fixes a bug where, on SEV-SNP, a certificate request could pass authorization with one os_image_hash while the issued certificate claimed another. Thats exactly where “attested” turns into “trust the cert”. On dstack? Upgrade guest OS, KMS and verifier. For context, NEAR’s GLM 5.3 Flash route shows up as Phala/dstack on its provider page and Redpill uses Phala attestation.

DDRop

DDRop is going to ACM CCS in November. Its a DDR5 interposer that silently drops memory writes. The attacker needs privileged control of the host plus brief physical access to install it, and the paper puts the parts at about $159 per system when building ten, not counting assembly, testing or R&D. It breaks memory integrity on TDX, Scalable SGX and SEV-SNP, and on TDX they forged attestation reports with it. Intel and AMD call it out of scope (of course).

The researchers expect NVIDIA’s confidential GPUs to be out of reach because HBM sits inside the chip package, but they mark that as inferred from documentation, not tested. So no, confidential-GPU services aren’t shown to be immune. Who physically operates the hardware still matters.

CC overhead keeps shrinking

Confidential AI published KV-cache offloading under CC on Oct 13. On B300s, CC on in both configurations, encrypted offloading did 141% more throughput than HBM-only serving on llm-d’s benchmark and cut mean time to first token from 39 s to 1.5 s. Careful tho, thats offloading vs no offloading. Its not a CC-on vs CC-off speedup.

If you want CC on vs off, NVIDIA’s own B200 numbers do that: on DeepSeek-R1 CC kept 96.1 to 98.2% of baseline output-token throughput (so 1.8 to 3.9% lost) and added 1.2 to 4.3% per-token latency. “TEEs are too slow” is getting hard to defend for inference.

New names (mostly routers)

Most new entrants are routers on top of a few enclave operators. Question is always the same: does the router verify the upstream enclave or just pass the claim along?

NanoGPT at least answers it honestly for Kev-4B, Jared Palmer’s decision model that Phala has started serving. Its model API says it “does not currently enforce attestation or verify signed receipts for this route”. So the enclave is Redpill’s claim, not something NanoGPT checks for you.

And one I don’t track yet and haven’t run my verifier against. Umbra went live on Oct 18, an x402 pay-per-request gateway paid in stablecoin. It says the gateway runs in an Intel TDX VM on Phala’s dstack. But its own domains, umbra-inference.app and api.umbra-inference.app, are on Vercel, so there TLS terminates on conventional cloud hardware before reaching the TDX gateway. Only the phala.network endpoint goes to the enclave directly. The “standard” tier is Claude via Anthropic, which sees the prompt (Umbra labels this). Calls itself a public testnet. None of this is verified.

Thats it for #1. Full issue and the data behind it are archived at confidentialinference.net, and attestation changes go out as they happen via the alerts RSS.

If you run one of these services and I got something wrong let me know.

Aljaz

https://fulcrumsec.vg/pingcap/ another day, another whoopsie

2026-10-21T07:40:44Z

https://fulcrumsec.vg/pingcap/ another day, another whoopsie

the book accelerando comes to mind frequently nowadays

2026-10-20T10:46:05Z

In reply to nevent1q…c6gq
_________________________

the book accelerando comes to mind frequently nowadays

Nostr event nevent1qqsqaxsd8mqdzksw0uqcfrc8xftks64eeusenv0a6u0lfyqydzw3yaqzyrh7t5fqmuxv9y86wjrj076943y8e2kngm20y2f6kp573uqlc5vcz2fkakd

2026-10-20T04:03:10Z

AGORISM.SPACE i was going over a list of domains that my past me ...

2026-10-19T14:46:16Z

AGORISM.SPACE

i was going over a list of domains that my past me bought when he thought that future me will have abundance of spare time and to give myself an excuse to keep it I had to build something

Agorism.space aggregates all the data from nostr marketplaces, has WoT so you can filter by your web of trust and other filters.
#agora #freemarkets #gamma OpenMarketsFoundation (nprofile…ycg7)

https://agorism.space/

You are not alone 😅

2026-10-18T20:06:09Z

In reply to nevent1q…s4rz
_________________________

You are not alone 😅

opens a foot fetish onlyfans account

2026-10-18T18:29:23Z

In reply to nevent1q…802c
_________________________

*opens a foot fetish onlyfans account*

#musicstr https://www.youtube.com/watch?v=3rFoGVkZ29w

2026-10-18T18:27:00Z

#musicstr https://www.youtube.com/watch?v=3rFoGVkZ29w

this dude is on fire, he made a DIY solar panel at home, also his ...

2026-10-18T09:16:13Z

this dude is on fire, he made a DIY solar panel at home, also his other stuff is worth checking out, truly a frontier of decentralized manufacturing at home https://www.semiconductor.diy/guides/microwave-solar-panels/

another day another #cybersecurity whoopsie ...

2026-10-18T04:59:34Z

another day another #cybersecurity whoopsie
https://quentinkaiser.be/security/2026/10/16/one-parser-two-bugs/

if you think confidential computing is hard, try naming things ...

2026-10-09T16:54:07Z

In reply to nevent1q…ku5g
_________________________

if you think confidential computing is hard, try naming things 🤣 🤣 😅

@nprofile…ze9l @nprofile…58km cap - confidential applications ...

2026-10-09T11:58:18Z

Gzuuus (nprofile…ze9l) Gigi (nprofile…58km) cap - confidential applications platform, paas - platform as a service 🥳

quoting
nevent1q…3g5y
38: Trusted Execution Environments and the Future of Human-Machine Interaction w/ gzuuus
https://haven.dergigi.com/6c44a491e36f1a2b8dbc62e44d9681755c5cfa63cacbf8bbb007bc81ca6af612.m4a
#NoSolutions

Nostr event nevent1qqs2lm5lv2x87e04fkpr2l44lnxgvatmw8dtl8vj939tzpafcjwp5sszyrh7t5fqmuxv9y86wjrj076943y8e2kngm20y2f6kp573uqlc5vcz96x3mh

2026-10-08T09:36:04Z

In reply to nevent1q…xee9
_________________________

Nostr event nevent1qqsgktvqy78wqufhykpszzu26zlrkvua7r5cewa7mcg2a2nvp9dsgrqzyrh7t5fqmuxv9y86wjrj076943y8e2kngm20y2f6kp573uqlc5vcz88x094

2026-10-07T19:10:14Z

false flag a day keeps the prosperity away

name one thing you trust more than the government... strippers ...

2026-10-06T17:36:51Z

name one thing you trust more than the government...

strippers who say they love me.

Nostr event nevent1qqsqqq9t9d022dl4gyavs4ayr7tlh34zz97cs2fg3w8uaj3wc34kayczyrh7t5fqmuxv9y86wjrj076943y8e2kngm20y2f6kp573uqlc5vcz6f9tyl

2026-10-04T10:43:09Z

#memes

The new KYC enforcement bot ...

2026-10-02T08:38:19Z

The new KYC enforcement bot https://www.youtube.com/watch?v=Ph63zHO9w7U

i'm happy to do it but I can only do remote

2026-10-02T08:10:14Z

In reply to nevent1q…h4z9
_________________________

i'm happy to do it but I can only do remote

Nostr event nevent1qqsx8l00tzf9dlrtvfm4t28n6z6dttvwyp2edcsvjdck0945rrgwk4czyrh7t5fqmuxv9y86wjrj076943y8e2kngm20y2f6kp573uqlc5vczhu8z47

2026-10-02T08:08:32Z

#memes

yeah 256gb is rough, i have one for a vibe box and i have to ...

2026-10-01T04:12:40Z

In reply to nevent1q…zfp7
_________________________

yeah 256gb is rough, i have one for a vibe box and i have to avoid doing any rust projects on it cuz the space just disappears

se par ekstra linkov o arku https://ark-protocol.com/

2026-09-30T18:13:54Z

In reply to nevent1q…5lyg
_________________________

se par ekstra linkov o arku https://ark-protocol.com/

#memes or not ...

2026-09-30T08:08:58Z

#memes or not

Governments are like condoms. A condom allows for inflation, ...

2026-09-29T12:01:45Z

Governments are like condoms. A condom allows for inflation, stops production, destroys the next generation, protects a bunch of dicks and gives you a sense of security while you are actually being screwed.

yes it does fall into private ai category. based on the claims (i ...

2026-09-29T10:39:24Z

In reply to nevent1q…mlca
_________________________

yes it does fall into private ai category.

based on the claims (i havent read the source) it provides anonymization of traffic to the providers, but that addresses a very specific issue - attribution of the conversation to specific identity.

it does not protect the privacy of the conversation itself, at least not from the provider. So if you upload sensitive documents or prompt it with releaving data your identity might not be tied to the account, but the data can still get out.

thats where the confidential inference and confidential computing part comes in, you can get much better assurances there provided that its implemented right and you can attest to a reproducible stack that is not leaking your data somewhere.

the weights you can verify if the env is fully reproducible the ...

2026-09-29T05:34:47Z

In reply to nevent1q…xclc
_________________________

the weights you can verify if the env is fully reproducible

the local attack vectors are a problem, ideally you have a secure data center + tamper proof hardware.

its not a panacea but theres a huge difference between "i'm using random llm api provider (or a proxy to have even more hops that you dont control)" and "i'm using a fully reproducible remotely attested confidential inference api" , even if the latter is not perfect for all threat models

Its also something that is defined differently by everyone ...

2026-09-28T19:56:09Z

In reply to nevent1q…u8wv
_________________________

Its also something that is defined differently by everyone depending on what they are selling

if anyone is running a gitlab instance somewhere you have to ...

2026-09-28T17:33:22Z

if anyone is running a gitlab instance somewhere you have to patch it https://docs.gitlab.com/releases/patches/patch-release-gitlab-19-3-2-released/

The spectrum of privacy in AI ...

2026-09-28T16:27:28Z

The spectrum of privacy in AI

Nostr event nevent1qqs8tr8rsnnr7s203wy7wyaukg4mz0mnyf3jdwfyjclgf8kcjttechszyrh7t5fqmuxv9y86wjrj076943y8e2kngm20y2f6kp573uqlc5vcz56lr9w

2026-09-28T04:28:56Z

#memes

the governments stealing from their people are not on nostr to ...

2026-09-25T14:33:21Z

In reply to nevent1q…9059
_________________________

the governments stealing from their people are not on nostr to tag

sadly no, the previous version didnt have export chat yet :(

2026-09-25T12:32:44Z

In reply to nevent1q…nxez
_________________________

sadly no, the previous version didnt have export chat yet :(

its been a long time coming..... NEW MANGO RELEASE (some) new ...

2026-09-24T14:13:42Z

its been a long time coming..... NEW MANGO RELEASE

(some) new stuff:
- direct onboarding to PPQdotAI (npub16g4…kv4h) (just pay with ln to get credits, no api key shenanigans)
- support for local/hybrid model use (if phone supports it), meaning you can have local model with confidential offloading
- discover and use remote tools over ContextVM (npub1dvm…3jdm)

the bad stuff:
- i had to generate a new cert to sign the app so you wont be able to upgrade, you'll have to do a clean install

https://zapstore.dev/apps/dev.disobey.mango

Nostr event nevent1qqsdhhv8yh5xvtfr4pkk37zv860em058nf6jds4q4hpmgeg3lrn6avczyrh7t5fqmuxv9y86wjrj076943y8e2kngm20y2f6kp573uqlc5vczunvxzd

2026-09-24T08:12:38Z

ouch https://radicle.dev/2026/09/23/disclosure-of-vulnerability-in-network-protocol

Its more something you'd have to compute from SBOM

2026-09-24T02:54:37Z

In reply to nevent1q…nhgw
_________________________

Its more something you'd have to compute from SBOM

the dream would be alerts on supply chain vulnerabilities for all ...

2026-09-23T15:05:31Z

In reply to nevent1q…f2qk
_________________________

the dream would be alerts on supply chain vulnerabilities for all apps

Nostr event nevent1qqswl239frw9lr86xx9smh6vdmt0pz3x0zptfmvmn0jk7v9jj3wcwhgzyrh7t5fqmuxv9y86wjrj076943y8e2kngm20y2f6kp573uqlc5vczakp7wz

2026-09-23T15:01:20Z

nothing gives you back more time than paying less taxes

random act of kindness ...

2026-09-22T07:25:46Z

random act of kindness

lots of bumfuckle going around ...

2026-09-22T07:14:14Z

lots of bumfuckle going around

It got a bit worse now since everyone started swaping put ssds ...

2026-09-19T14:50:55Z

In reply to nevent1q…u2dp
_________________________

It got a bit worse now since everyone started swaping put ssds for smaller ones and everything has less ram all of the sudden, but generally known reputable providers who refurbish fleets of large companies tend to be ok, tho my exp is all from europe

cars can be outside, servers cant.

2026-09-18T03:18:25Z

In reply to nevent1q…p58q
_________________________

cars can be outside, servers cant.

"Google Pixel devices contain an improper authorization ...

2026-09-17T19:26:22Z

"Google Pixel devices contain an improper authorization vulnerability in the cellular modem," CISA said. "A logic error may allow an attacker to bypass permission checks and escalate privileges."

https://thehackernews.com/2026/09/google-patches-pixel-modem-flaw-amid.html

State of AI #meme ...

2026-09-17T08:17:22Z

State of AI #meme

theres also the reddit one who got a lot ...

2026-09-16T08:27:00Z

In reply to nevent1q…rvzl
_________________________

theres also the reddit one who got a lot more

Nostr event nevent1qqsdx2tgn6zjs8yuj0ny0tcqg2w0ktf0dpe8ghqv076qlsgltfdc6aszyrh7t5fqmuxv9y86wjrj076943y8e2kngm20y2f6kp573uqlc5vczv75jwd

2026-09-16T05:42:19Z

#memes a

Good fucking monday! ...

2026-09-14T06:21:15Z

Good fucking monday!

one of the servers running nostr.at/eu/ae njump gateways ran out ...

2026-09-12T15:11:24Z

one of the servers running nostr.at/eu/ae njump gateways ran out of monthly traffic allowance (1.5TB) so expect some more timeouts in the next days (or zap heavily for more infrastructure :D)

how it feels to be a bitcoiner in 2026 ...

2026-09-11T17:28:23Z

how it feels to be a bitcoiner in 2026

Danger zone

2026-09-10T20:37:22Z

In reply to nevent1q…d8h4
_________________________

Danger zone

Nostr event nevent1qqsg8gj4aqtm87l9td28pfrkmpm24m9v97h07pmp9r6t5k5d20r26zgzyrh7t5fqmuxv9y86wjrj076943y8e2kngm20y2f6kp573uqlc5vczy3ugtp

2026-09-09T05:58:08Z

when was the last time you pasted your nsec in a new app?

at least this timeline is producing good memes ...

2026-09-08T08:07:50Z

at least this timeline is producing good memes

maybe law of salvage should be applied here in regards to the ...

2026-09-07T14:48:01Z

In reply to nevent1q…ncgr
_________________________

maybe law of salvage should be applied here in regards to the bounty

ewww, undies

2026-09-07T05:06:35Z

In reply to nevent1q…vrjv
_________________________

ewww, undies

I've been in bitcoin for a decade and I've only now ...

2026-09-06T06:26:51Z

I've been in bitcoin for a decade and I've only now started to mess with mining - used glm-5.3 and astra to make improvements to SparkMiner firmware and squeeze 13% more performance out of my esp32 based nerdminer board

what does the fox say #musicstr ...

2026-09-03T18:02:54Z

what does the fox say #musicstr https://www.youtube.com/watch?v=jofNR_WkoCE

"bitcoiners" love telling others what to do to meet their ...

2026-09-03T09:58:04Z

"bitcoiners" love telling others what to do to meet their purity standards. such freedom maxing.

the fact that confidential inference is in not the greatest of ...

2026-09-02T18:59:45Z

In reply to nevent1q…qle9
_________________________

the fact that confidential inference is in not the greatest of states around as seen in results at https://confidentialinference.net/attestation is one of the main reasons i'm building and sdk (still in early stage) https://github.com/enclava-labs/confidential-inference-sdk

the idea is to have a no nonsense way to integrate confidential inference properly, not just connecting to their openai compatible endpoint and praying for the best. this will enable you to fail close if attestation fails, it will hopefully also kick all providers asses to start doing things properly

the problem a lot of times is not necessarily that anyone is malicious, its just tedious to always have everything up2date and even more importantly we lack approaches that would be adopted by multiple providers so everyone rolls their setup a bit differently which leads to different attestation approach, different ways of communicating the versions etc.

also lets not get started about how easily reproducible the envs are, thats gonna be another story

for all of you poor souls that were born in europe at the right ...

2026-09-02T15:21:22Z

for all of you poor souls that were born in europe at the right time to be exposed to kelly family #musicstr https://www.youtube.com/watch?v=e1xkb9y34YA

i think you described the timeline, not the potential outcomes

2026-09-02T06:51:29Z

In reply to nevent1q…42mh
_________________________

i think you described the timeline, not the potential outcomes

asking the real questions #meme ...

2026-09-02T06:50:35Z

asking the real questions #meme

you know its gonna cost a kidney or two when theres no pricing on ...

2026-08-27T05:53:31Z

In reply to nevent1q…9743
_________________________

you know its gonna cost a kidney or two when theres no pricing on the website

pretty good read about TEE security at signal app and how it got ...

2026-08-26T19:00:47Z

pretty good read about TEE security at signal app and how it got exploited https://v12.sh/blog/signal

looking for an mp3 of: A Mind at Play - How Claude Shannon ...

2026-08-26T12:39:34Z

looking for an mp3 of: A Mind at Play - How Claude Shannon Invented the Information Age by Jimmy Soni and Rob Goodman, sats reward available

for any #reticulum enthusiasts i'm working on full feature ...

2026-08-26T11:32:33Z

for any #reticulum enthusiasts i'm working on full feature parity between the beechat's rust sdk and the official python sdk, if anyone wants to help testing things or look at it any feedback is welcome https://github.com/aljazceru/Reticulum-rs

AI #meme ...

2026-08-25T08:33:14Z

AI #meme

Truth is funny #meme ...

2026-08-22T19:52:39Z

Truth is funny #meme

Life of a vibe coder ...

2026-08-20T08:12:26Z

Life of a vibe coder

state of the world ...

2026-08-18T15:22:00Z

state of the world

a look into beloved hetzner datacenter ...

2026-08-18T12:46:14Z

a look into beloved hetzner datacenter
https://www.youtube.com/watch?v=MQeJYEN_lrg

i've done finance and energy grids (among other things) ...

2026-08-18T06:54:23Z

In reply to nevent1q…6fpq
_________________________

i've done finance and energy grids (among other things) before, freedom tech in general is very odd in a sense that you go from everything is mission critical to most things are broken most of the time

runs

2026-08-18T06:52:02Z

In reply to nevent1q…qa02
_________________________

runs

i do - i blocked them

2026-08-17T20:00:41Z

In reply to nevent1q…3624
_________________________

i do - i blocked them

somewhat reminds me of https://craphound.com/category/walkaway/

2026-08-17T18:58:23Z

In reply to nevent1q…hays
_________________________

somewhat reminds me of https://craphound.com/category/walkaway/

tons of traffic from https://lightpanda.io/ it appears

2026-08-17T18:37:28Z

In reply to nevent1q…nudw
_________________________

tons of traffic from https://lightpanda.io/ it appears

and people say theres no users on nostr 😹 ...

2026-08-17T18:25:58Z

and people say theres no users on nostr 😹

Nostr event nevent1qqsvz45zcknkp7afpwnly94yefqsh8kacp5luzdy2lsf6c9j6540gvszyrh7t5fqmuxv9y86wjrj076943y8e2kngm20y2f6kp573uqlc5vczk7pyss

2026-08-17T11:53:08Z

In reply to nevent1q…zyxk
_________________________

@nprofile…td2a i think you're having some issues with ...

2026-08-17T09:24:40Z

sondreb (nprofile…td2a) i think you're having some issues with nostria, i'm seeing host timeouts from cloudflare

initially i thought you're using machetes not feathers as ...

2026-08-16T19:28:12Z

In reply to nevent1q…3jhq
_________________________

initially i thought you're using machetes not feathers as props 😅

#musicstr https://www.youtube.com/watch?v=2uTMTyqQxl4

2026-08-16T19:12:24Z

#musicstr https://www.youtube.com/watch?v=2uTMTyqQxl4

Nostr event nevent1qqswr60fhus6gewm9gc4tf7xx380tk75p54f0nnxq6cj5nppyurr96qzyrh7t5fqmuxv9y86wjrj076943y8e2kngm20y2f6kp573uqlc5vczaek8cc

2026-08-16T08:45:46Z

Nostr event nevent1qqs9vn8m8kevae3q826wtry0suk90qhpqe74x82t0lfhpv6c24hcxuszyrh7t5fqmuxv9y86wjrj076943y8e2kngm20y2f6kp573uqlc5vczghmfdk

2026-08-08T18:07:15Z

In reply to nevent1q…u44d
_________________________

RNG as fake as the tits 😂

2026-08-08T14:01:45Z

In reply to nevent1q…fzn0
_________________________

RNG as fake as the tits 😂

Nostr event nevent1qqsz5524t0hh7tpjz8yjm62ylarh24699yp6080dthwqeh2qu7cxq4gzyrh7t5fqmuxv9y86wjrj076943y8e2kngm20y2f6kp573uqlc5vcza2yllt

2026-08-08T05:18:07Z

quoting
nevent1q…vfw0
we don't disclose which projects we scan, or the results

Nostr event nevent1qqsp9pea5rj30dhjwdq4mzmlfn7xp6k88e0c29nzpas6y0qnw3w0lgszyrh7t5fqmuxv9y86wjrj076943y8e2kngm20y2f6kp573uqlc5vcz4zx5q6

2026-08-08T05:06:34Z

In reply to nevent1q…z92c
_________________________

nostr://nevent1qqsqqqrn0q0up2lmt2ffvujyth2fy4nnptljt3lxdhej7nlgcqqmakq3p6zxk

i have a website which provides more info about providers and ...

2026-08-07T13:53:01Z

In reply to nevent1q…7krx
_________________________

i have a website which provides more info about providers and their reliability (i've removed mape because they require subscription to even list the models they support and their token pricing is not publicly available - only an old documentation links)

most companies in this space just resell someone elses infrastructure for a premium and prices tend to vary widely. same with their actual correctness of confidential infrastructure implementation (reproducability , attestation etc)

https://confidentialinference.net/

i didn't say doing AI reviews of the projects makes sense, ...

2026-08-07T12:27:44Z

In reply to nevent1q…jysv
_________________________

i didn't say doing AI reviews of the projects makes sense, what I question is the need to spend all the effort in announcing the great successes of the beloved red team

not to mention the incredibly sloppy instances of public msgs across platforms tagging founders/maintainers saying "hey read your DMs we need to discuss something security related".

doing the work and making sure everyone knows you are doing the work are two different things.

incentives and expected value calculations are a bit different when you as an attacker need to guess/hope that some of the projects will bare fruit of your attacks vs having confirmation that theres tons of vulns so just dig and you'll find more gold. specially because theres no value in spreading the stats of the results besides telling everyone how great of a job you are doing.

i think you missed some

2026-08-06T14:24:19Z

In reply to nevent1q…xrn2
_________________________

i think you missed some

worse case would be that this speeds up someone looking into ...

2026-08-06T14:23:27Z

In reply to nevent1q…herh
_________________________

worse case would be that this speeds up someone looking into something very commonly deployed and another fuckton of ppl lose their money cuz of it

Can we maybe stop screaming how many critical vulnerabilities ...

2026-08-06T14:06:37Z

Can we maybe stop screaming how many critical vulnerabilities were found in all the projects that are being scanned are found, this is pure insanity.

Why are you painting an even bigger target on all the projects? Scan the repos, contact the maintainers privately and stfu about it so not everyone and their mom gets curious since you are so loudly proclaiming tons of vulns

Everyone is always so opinionated and loud about responsible disclosure but now that theres a "bitcoin red team" we are all screaming about critical vulns they keep discovering?

Nostr event nevent1qqs8wpluxt40tq30a8shg7y6v9hqk3ylg5z7supxzjnx5gfa4qtrudczyrh7t5fqmuxv9y86wjrj076943y8e2kngm20y2f6kp573uqlc5vczkuuj9c

2026-08-05T09:47:30Z

😂

Nostr event nevent1qqsz2920gr8lzeusztd7rg7zrxegduunvur4gpr2fusacjauq5rfwcszyrh7t5fqmuxv9y86wjrj076943y8e2kngm20y2f6kp573uqlc5vcz623qg5

2026-08-05T08:03:01Z

iykyk

i must resist another distraction - open source vacuum robot ...

2026-08-05T07:53:56Z

i must resist another distraction - open source vacuum robot https://github.com/makerspet/oomwoo