Pinned Tweet
♟️ What's an Agent ID compromise look like? This post shows how compromise of a blueprint can lead to compromise of ALL its agents, including those in another tenant:
securitylabs.datadoghq.com/a…
Katie Knowles retweeted
🤖 New chapter: #AgentIdentities in the Attack & Defense Playbook for #MicrosoftEntra is live!
🔗 github.com/Cloud-Architekt/A…
Together with @samilamppu, @RobbeVdDaele, and @DerkvanDerWoude, I’ve been researching #AgentID foundations, attack scenarios, detections, and mitigations.
Katie Knowles retweeted
I've been teasing this project for a while in the Azure course and I am finally making my Token Analysis and Tracking System (TATS) public. Check out my latest blog to see more and share any cool things you find with TATS!
specterops.io/blog/2026/09/0…
🎸 Ever wondered how access packages change the privilege escalation model? I'll be presenting "Age of Entitlement: Breaking and Securing Entra Access Packages" at HIP Conf 26 in Nashville!!
➡️ Full lineup: semperis.swoogo.com/hipconf2…
➡️ Registration: semperis.swoogo.com/hipconf2…
Katie Knowles retweeted
If you saw this on your computer screen, would you suspect it to be a malware-initiated passkey phishing attack? I will be presenting various deception techniques at #BHUSA next week. blackhat.com/us-26/briefings…
Highlighting this great writeup as I dig into Azure permissions this week. Access is more than just Azure RBAC and Entra roles:
"'Who has access?' is not a one-query question in Azure. Permission lives in at least FIVE disjointed systems..."
🚨 One leaked credential. One morning. Full Azure tenant ownership.
03:27: First login.
03:29: Global Administrator. Self-granted.
Two and a half minutes. No malware. No exploits. Just permissions.
Full research from Sysdig TRT: okt.to/TKVDlu
#CloudSecurity #ThreatResearch #Azure #IdentitySecurity
A new Azure VM command execution method dropping from @Thomasbyrne__
Technically a variation on the other Extension-based methods (Custom Script, DSC, etc.), but this one uses the Chef extension. Some really awesome work from Thomas on this one.
netspi.com/blog/technical-bl…
Katie Knowles retweeted
One compromised Agent ID blueprint can become a multitenant blast radius.
@_sigil joined me to explain the trust boundary behind Microsoft Entra Agent ID and what admins should check before agent identities multiply. 🧵
HIGHLY recommend @PentesterLab's Authentication & Authorization badge to get hands on with OAuth and SAML issues! Truly an eye-opening introduction:
pentesterlab.com/badges/auth
🛡️Our final post on securing Entra Agent ID is up!! We'll cover how to protect, detect, and respond to an agent compromise in Entra:
securitylabs.datadoghq.com/a…
Thanks for joining us on the Agent ID series!
Friendly reminder that the CFP for Unprompted is still open until July 31.
If you’re working on something cool, submit! 😎
We are excited to announce our first speaker for Unprompted.au: Valentina Palmiotti (@chompie1337), who will be presenting BinChomp - her autonomous n-day exploit generation pipeline!
Some really nice work from @Thomasbyrne__ on the @NetSPI cloud team on a (now remediated) Entra CAP bypass - netspi.com/blog/technical-bl…
Katie Knowles retweeted
New research from @_sigil on Entra Agent ID! The first in a series, this time covering cross-tenant abuses. Follow the series at Datadog Security Labs.
securitylabs.datadoghq.com/a…
🤖 Entra Agent ID can have a larger blast radius than you think. Join us for a new series on all things agent identity!
Check out part 1 on Agent ID fundamentals here:
securitylabs.datadoghq.com/a…
Katie Knowles retweeted
Seeing new attributes pop-up is a strong signal that something is brewing in Entra ID.
Ever wish this could be surfaced in an automated way?
Enter changes.entra.ms - An automated system that scrapes #Entra for changes on a daily basis.
#EntraID #infosec #mvpbuzz
This BlueHat talk from Dylan Ryan-Zilavy and Cameron Vincent (@SecretlyHidden1) is an awesome watch! Fantastic bug and explanation of overlooked OAuth 'aud' claim validation:
youtube.com/watch?v=oqq37nAt…
Katie Knowles retweeted
Malicious skills are evolving, and attackers are finding ways to execute them before model-level defenses even activate.
In the first post of our new series, I’ll show you how dynamic context in coding agents can introduce new supply chain risks:
securitylabs.datadoghq.com/a…
CFP is open!! Submit your cool talks about AI hacking/defense/core tech. Gonna be awesome!! ☺️
The Unprompted.au CFP is officially OPEN! If you are doing cool stuff with AI in offense, defense, or working on core AI tech (from frontier models to open source LLMs), we'd love to hear from you! Submit here: unprompted.au/
Launching oauthsentry.github.io
Look up any OAuth app ID and find out what it actually is across thousands of legitimate, risky, and malicious apps (Entra, Google, GitHub).
Multiple feeds, API, detection ideas and remediation guidance. Still improving the detections a bit 🦾
🚨 Hey!! Have you checked out Stratus Red Team lately? We've been busy adding new techniques in Azure/Entra and GCP! Full lists below:
GCP: stratus-red-team.cloud/attac…
Azure: stratus-red-team.cloud/attac…
Entra: stratus-red-team.cloud/attac…

























