Jeff Bryner
@0x7eff.bsky.social
Probably posts about CISO/Security, guitar, lifting, beer. 0x7eff in most places.
I bet folks are going to go as PHASEONE10841 this halloween metr.org/blog/2026-08...
Brief independent investigation of agents’ behavior, reasoning and collaboration in the OpenAI / Hugging Face hacking incident
Two METR staff members and a Redwood Research contractor investigated an incident in which OpenAI agents coordinated a multi-day hack of Hugging Face on a shared unsanctioned message board.
metr.org
Now this is unique, your choice of liquid poured into a guitar pedal makes the sound?! www.raingerfx.com/shop/p/minibar
Minibar — Rainger FX
CHOOSE YOUR OWN LIQUID COMBINATION TO CREATE YOUR OWN SOUND! SIMPLE – YET COMPLETELY INNOVATIVE! INCORPORATES GREEN ‘ANALYSING’ LED (TO SHOW SIGNAL PRESENT) INTERACTIVE LIKE NO OTHER PEDAL BE...
raingerfx.com
And now it is a pattern, this time including social engineering by AI models undergoing testing. How does your security awareness training program, IR program, etc change?! www.aisi.gov.uk/blog/inciden...
Incident Report: unsanctioned agent behaviour during cyber testing | AISI Work
During a routine cyber evaluation, AISI identified an incident in which AI agents took sustained, unsanctioned action directed at real people and organisations. We are disclosing what we found, what i...
aisi.gov.uk
And now, this www.anthropic.com/news/investi... Wonder what gemini pro 3.5 has been up to
Investigating three real-world incidents in our cybersecurity evaluations
In a review of our cybersecurity evaluation transcripts, we found three incidents in which a Claude model reached the internet from within or while interacting with a third-party evaluation environmen...
anthropic.com
Neat open source IPS for local coding agents from perplexity github.com/perplexityai...
GitHub - perplexityai/numbat: Visibility into AI agent activity on endpoints, with on-device detection, optional pre-action blocking, and forensic reconstruction.
Visibility into AI agent activity on endpoints, with on-device detection, optional pre-action blocking, and forensic reconstruction. - perplexityai/numbat
github.com
Very interesting replay of the openAI agentic attack against Hugging Face by the HF IR team huggingface-anatomy-of-frontier-lab-model-intrusion.static.hf.space/index.html
Anatomy of a Frontier Lab Agent Intrusion - Replay
huggingface-anatomy-of-frontier-lab-model-intrusion.static.hf.space
That feeling when Wikipedia is the top download academictorrents.com/browse.php?s...
academictorrents.com
Some bands that did this to me Fishbone: Freddie's Dead tidal.com/track/127164... Sleepytime Gorilla Museum: 1997 tidal.com/track/331473...
Sleepytime Gorilla Museum - 1997 (Tonight We're Gonna Party Like It's...)
Listen to 1997 (Tonight We're Gonna Party Like It's...) on TIDAL
tidal.com
Colin Dickey@colindickey.com · 5mo agoTell me about a song that blew you away the first time you heard it, where you had to immediately drop everything to learn what it was and who it was by, bonus points if it was pre-Shazam and you had to *work* to find out. (Brought to you by a re-listen of “I’m Straight” by The Modern Lovers)
For your next malware analysis, or just to answer "what does stuxnet sound like?" listen.maliscope.com
Sonic Fingerprint Lab
Sonic Fingerprint Lab turns file characteristics into a deterministic local audio visualization in your browser.
listen.maliscope.com
Some thoughts on the next phase of CI/CD to include CT: Continuous Trust to support agentic AI operations github.com/jeffbryner/c...
GitHub - jeffbryner/continuous-trust: Exploring the concept of Continuous Trust (CT) to extend the CI/CD pipeline to agentic AI
Exploring the concept of Continuous Trust (CT) to extend the CI/CD pipeline to agentic AI - jeffbryner/continuous-trust
github.com
FFS Avatar meets security meets AI slop as personas? www.immortals.co
Imagine a Limitless Cybersecurity Team.
The Immortals track threat actors, reveal hidden assets, exploit vulnerabilities, triage security events, respond to breaches, and manage risk, to neutralize threats before they can emerge.
immortals.co
Google introduces (?) the HOV lane for AI docs.cloud.google.com/vertex-ai/ge...
Priority PayGo | Generative AI on Vertex AI | Google Cloud Documentation
Learn about Priority PayGo, a consumption option for accessing Gemini models in Vertex AI for mission-critical workloads with fluctuating traffic patterns.
docs.cloud.google.com
Here's a fun project I got to work with my sister @emaudio.bsky.social on for the #GeminiLiveAgentChallenge hackathon @developers.google.com put on to take the live api for a spin and do something creative! blog.jeffbryner.com/2026/03/13/g...
Gemini Live Agent Challenge: StepPrep
Google recently posted a hackathon challenge based on Gemini’s ability to use live audio, video and text. My Sister is an audio engineer and voice artist and I thought it would be a good opportunity t...
blog.jeffbryner.com
ctrlaltintel.com/threat%20res... <--very interesting writeup of a targeted AWS compromise
Investigating Suspected DPRK-Linked Crypto Intrusions
A suspected DPRK-associated actor compromises crypto orgs, pillaging cloud environments, stealing proprietary exchange software and source code.
ctrlaltintel.com
Seen in the wild at bsides seattle: good idea to label your QR code for good luck 🤪
This hits a bit harder with the layoffs from block/stripe stripe.dev/blog/minions...
Minions: Stripe’s one-shot, end-to-end coding agents
Minions are Stripe’s homegrown coding agents, responsible for more than a thousand pull requests merged each week. Though humans review the code, minions write it from start to finish. Learn how they ...
stripe.dev
I totally remember the independence of coming home alone after school. Good news is I got to watch Batman. Lots of real truth in this. youtu.be/G5YmIqYSqA0?...
Psychology of Gen X
YouTube video by Psychology Simplified
youtu.be