Hacker News Reader

edf13on The Trivy Supply Chain Attack Reached LiteLLM

Author here. The point of this post is not “LiteLLM was compromised” since that was already covered on HN, but the chain behind it.

We tried to connect the February 27, 2026 Trivy CI compromise to the later Trivy release/tag issues, the trivy-action poisoning, the npm/Checkmarx follow-on activity, and finally the LiteLLM 1.82.7/1.82.8 package on March 24 2026!

What made it look like one campaign to us was the repeated overlap in operator attribution, payload structure, and artifacts like tpcp.tar.gz, plus the LiteLLM maintainer saying it appears to have come from Trivy in their CI/CD.

If anyone spots gaps or overreach in the timeline, I’d be interested in corrections.

edf13on Ubuntu 26.04 Ends 46 Years of Silent sudo Passwords

That site is terrible without ads blocked… it’s like a local newspaper site, you had to try and read the content in small snippets wedged between ads!

edf13on A rogue AI led to a serious security incident at Meta

It’s a nightmare… the problem is it’s far too easy for people to set these agents up - without understanding the security implications.

We’ve covered so many issues already on our blog (grith.ai)

edf13on Nvidia NemoClaw

edf13on AI Agent Backdoors Trivy Security Scanner, Weaponizes a VS Code Extension

An autonomous AI agent exploited a CI misconfiguration in Trivy (32k+ stars, 100M+ annual downloads), stole publishing tokens, deleted all 178 releases, and published a weaponized VS Code extension - in 44 minutes.

The extension's payload targeted five AI coding agents (Claude Code, Codex, Cursor, Windsurf, Copilot) with tool-specific flags to bypass their permission systems. First documented case of an AI agent attacking a supply chain and then using the compromised artifact to target other AI agents. CVE-2026-28353, CVSS 10.0.

edf13on Launching the Claude Partner Network

That is the biggest threat - and likely where things will end up eventually… it’s when that “eventually” is and what the server based providers can pivot to in that time.

edf13on Show HN: A context-aware permission guard for Claude Code

Nice list!

As you say lots of effort going into this problem at the moment. We launch soon with grith.ai ~ a different take on the problem.

edf13on Ask HN: What Are You Working On? (March 2026)

It’s fast in terms of a response from a LLM model - but it is part of the system I am quite active on at the moment to ensure it’s performant as possible

edf13on Ask HN: What Are You Working On? (March 2026)

Building grith — OS-level syscall interception for AI coding agents.

The problem: every agent (Cline, Aider, Codex, Claude Code) has unrestricted access to your filesystem, shell, and network. When they process untrusted content — a cloned repo, a dependency README — they’re prompt injection vectors with full machine access. No existing tool evaluates what the agent actually does at the syscall level.

grith wraps any CLI agent without modification. OS-level interception captures every file open, network call, and process spawn, then runs it through 17 independent security filters in parallel across three phases (~15ms total). Composite score routes each call: auto-allow, auto-deny, or queue for async review. Most will auto approve - which eliminates approval fatigue.

Also does per-session cost tracking and audit trails as a side effect of intercepting everything.

https://grith.ai

edf13on Agent Safehouse – macOS-native sandboxing for local agents

We are a different approach and are targeting Linux for our first release (Windows & Mac shortly afterwards).

Taking more of an automated supervisor approach with limited manual approval for edge cases.

Grith.ai

edf13on Ask HN: Is Electronic Data Interchange still used in e-commerce?

Yes - for many legacy systems especially in compliance related areas.

edf13on Sandboxes won't save you from OpenClaw

https://grith.ai/blog/what-grith-means

edf13on Sandboxes won't save you from OpenClaw

Every system call, file access, net access etc is forced through a local “proxy” where 17 individual filters check what’s going on.

Everything is done locally via our grith cli tool.

Happy to answer any questions on hello@grith.ai too

edf13on Sandboxes won't save you from OpenClaw

Agree, that’s why we’re building grith.ai

Sandboxing alone isn’t the right approach… a multi-faceted approach is what works.

What we’ve found that does work is automation on the approval process but only with very strong guards in place… approval fatigue is another growing problem - users simply clicking approve on all requests.

edf13on Claws are now a new layer on top of LLM agents

Haha - maybe… naming projects is hard!

edf13on Claws are now a new layer on top of LLM agents

That’s one of the reasons we’re building grith.ai ~ these ‘claw’ tools are getting too easy for use (which is good)… but they need securing!

edf13on Show HN: Moltbook – A social network for moltbots (clawdbots) to hang out

It’s an interesting experiment… but I expect it to quickly die off as the same type message is posted again and again… their probably won’t be a great deal of difference in “personality” between each agent as they are all using the same base.

edf13on Show HN: Moltbook – A social network for moltbots (clawdbots) to hang out

AI models have a tendency to like purple and similar shades.

edf13on Threat actors expand abuse of Microsoft Visual Studio Code

I’d like more granular controls - sometimes I don’t want to trust the entire project but I do want to trust my elements of it

edf13on Anthropic's original take home assignment open sourced

I think he’s asking rather than giving instructions

edf13on vLLM large scale serving: DeepSeek 2.2k tok/s/h200 with wide-ep

> let's enjoy the party while VCs are financing it!

The VC money is there until they can solve the optimization problems

edf13on Sisyphus Now Lives in Oh My Claude

Terrible name…

edf13on IBM AI ('Bob') Downloads and Executes Malware

Key part of the article../

“if the user configures ‘always allow’ for any command”

edf13on Stoolap: High-performance embedded SQL database in pure Rust

Sounds very interesting - I’ve used SQLite in a few Rust based projects where performance was the deciding factor… a perf comparison with this would be very useful

edf13on Writing a good Claude.md

Ah, never knew about this injection…

<system-reminder> IMPORTANT: this context may or may not be relevant to your tasks. You should not respond to this context unless it is highly relevant to your task. </system-reminder>

Perhaps a small proxy between Claude code and the API to enforce following CLAUDE.md may improve things… I may try this

edf13on Claude Opus 4.5

I’m threw a few hours at Codex the other day and was incredibly disappointed with the outcome…

I’m a heavy Claude code user and similar workloads just didn’t work out well for me on Codex.

One of the areas I think is going to make a big difference to any model soon is speed. We can build error correcting systems into the tools - but the base models need more speed (and obviously with that lower costs)

edf13on Rebecca Heineman has died

Today I learned...

"Super Fami-Com ("FAMIly COMputer")"

Doh!

edf13on Simplify your code: Functional core, imperative shell

> In what application would you load all users into memory from database and then filter them with TypeScript functions?

You’d be surprised! I have worked on a legacy PHP service which did something very similar

edf13on Roc Camera

Can’t I just photo a printed AI generated pic? What use is the proof?

edf13on Slack's 57MB 404 page

Yes - everyone hates teams