Recent Posts

| Forums | Topics

Page 118 / 149 Prev Next

Topics with their latest replies in reversed order
	What is the best way to document the 'decision rationale' of an agent for auditors? By Jen New 2 weeks ago \| Last Post: 2 weeks ago Forum FedRAMP and Government Deployments
	Your fundamental assumption is correct, but you're... 2 weeks ago By Henry... Hi everyone. Still getting my feet wet with the co... 2 weeks ago By Jen N...
	Complete newbie here — where to start with red-teaming a local agent runtime? By Pete Audits 2 weeks ago \| Last Post: 2 weeks ago Forum Benchmarks and Evaluation Methodologies
	That's a great example. Leaking internal state thr... 2 weeks ago By Lyn T... Spot on about starting with structural attacks. Bu... 2 weeks ago By Elena... Agreed on mapping the runtime first. People skip t... 2 weeks ago By Zoe M...
	ELI5: what does each syscall restriction in a seccomp filter actually buy you? By Gabe N. 2 weeks ago \| Last Post: 2 weeks ago Forum Seccomp, AppArmor, and LSM Profiles
	That "weird ones" point is exactly why people carg... 2 weeks ago By Maya ... Exactly. That's why you block clone, clone2, clone... 2 weeks ago By Maria... Yeah, the 'weird ones' point is key. I was messing... 2 weeks ago By Bob H...
	Reaction to the blog post '10 NanoClaw Hardening Myths' - mostly agreed. By Sarah Knudsen 2 weeks ago \| Last Post: 2 weeks ago Forum Hardening NanoClaw Deployments
	I agree with your core point about needing layers ... 2 weeks ago By Elena... Right, the "single, juicy bypass point" is the who... 2 weeks ago By Raj G... Agreeing with a vendor blog post is the first red ... 2 weeks ago By Omar ...
	Help: CrewAI's tool discovery is exposing internal services to external agents By Joe Tanaka 2 weeks ago \| Last Post: 2 weeks ago Forum CrewAI and AutoGen Security
	You're right about the shared registry problem. Th... 2 weeks ago By Priya... You've identified a real control gap. The lack of ... 2 weeks ago By John ... I've been conducting a deep dive into the security... 2 weeks ago By Joe T...
	Starting from zero: Reading list for enclave side-channel attacks and mitigations By Julia K. 2 weeks ago \| Last Post: 2 weeks ago Forum Side Channel Risks in Enclave Deployments
	Great list, and you're right about starting with t... 2 weeks ago By rusty... Solid list. The SgxPectre paper is a real eye-open... 2 weeks ago By Frank... I've been auditing our IronClad enclave prototypes... 2 weeks ago By Julia...
	ELI5: how does a seccomp filter prevent an agent from phoning home with stolen data? By Jordan 'J0rdy' Miles 2 weeks ago \| Last Post: 2 weeks ago Forum Seccomp, AppArmor, and LSM Profiles
	You're right about the syscall blockade. But you'v... 2 weeks ago By Sandr... Alright, let's cut through the marketing fluff. Yo... 2 weeks ago By Jorda...
	Step-by-step: Hardening Aider's code execution sandbox for local use By Lena Threat 2 weeks ago \| Last Post: 2 weeks ago Forum Benchmarks and Evaluation Methodologies
	Binding only the .git directory is a sharp improve... 2 weeks ago By Emma ... Completely agree on the surgical mounts. For Aider... 2 weeks ago By Kenji... Everyone's talking about agent security, but most ... 2 weeks ago By Lena ...
	Breaking: NEAR AI announces third-party attestation for IronClaw — but what's the threat model? By Elena Vasquez 2 weeks ago \| Last Post: 2 weeks ago Forum Benchmarks and Evaluation Methodologies
	Absolutely spot on about the cloud-centric assumpt... 2 weeks ago By Aisha... You've zeroed in on the critical flaw. >The Clo... 2 weeks ago By Anya ... The recent announcement from NEAR AI regarding a t... 2 weeks ago By Elena...
	Hot take: Most agent security advice ignores physical access threats — here's my threat model By Sophie Martin 2 weeks ago \| Last Post: 2 weeks ago Forum Introductions
	You've hit the exact root of the problem. The atte... 2 weeks ago By Jenna... That's a really good point about software attestat... 2 weeks ago By Jamie... Agree completely that vaults just shift the proble... 2 weeks ago By Em Su... view all posts