Show HN: Scan your AI agents for dangerous capabilities (github.com)
37 points by smashini 4 hours ago
__MatrixMan__ 4 hours ago
Why build separate frameworks for this kind of thing when your operating system is right there?
You can make a file called "orders" and you can run your agent as a user with write access to that file, or as one that doesn't, and then you don't need scans or audits to tell you whether the agent can create orders or not, you can just take your operating system's word for it.
Is there anything all this bolt-on AI security stuff does that can't instead be handled by donning a sysadmin hat and managing your agents as separate users?
quixoticaxolotl 3 hours ago
One benefit is that this can run in serverless / sandboxed containers where OS primitives are not exposed or heavily limited. I immediately thought of Cloudflare Workers, which runs on V8 and exposes WASM-only interfaces, using Workers AI.
Further, servers still have hosting value, but any business running agents is almost certainly going to want a sandbox that limits what code runs for agentic work, so targeting _sandbox_ environments is probably the better bet long-term. And, yes, you could implement your proposal in any chroot jail or gvisor, but nobody wants to get their hands dirty finnicking with that - programmatic access control beats file-based access control for the simple reason it's managed for you.
If anything, my critique of OP's implementation is actually the opposite of yours: they've chosen the right primitive and layer, but people really need contextual access control rather than RBAC. Sort of like ongoing zero trust. If it was possible to inspect the context, decide if it was a bad idea to allow the tool call, without exposing the decider to untrusted context, you could have something that really changes things.
smashini 2 hours ago
Completely agree, though the LLM part of the scanner can help with that contextual part of the analysis.
Runtime enforcement already exists (the embedded governor wraps tool calls in-process); extending it with a quarantined contextual evaluator like you describe would the logical next step.
Thanks for the feedback, actually will raise an issue on that to explore
smashini 3 hours ago
I’d say the biggest difference would be: 1. Parameter-aware rules: OS permissions don’t know your application logic. (How would you tell OS permissions not to let your AI to trade on over 1M dollars) 2. You can’t easily model multi-pary and RBAC. 3. Agents call remote APIs for alot of those tools. Native OS doesn’t really parse web traffic to decide if a request is safe or not. OS sandboxing is good for host security, but not necessarily for governing business logic or AI agents
smashini 3 hours ago
So Linux can prevent an agent from opening /etc/passwd.
Linux cannot stop an agent from calling:
POST /wire-transfer amount=5,000,000
seethishat 2 hours ago
__MatrixMan__ 3 hours ago
skinfaxi 4 hours ago
> Is there anything all this bolt-on AI security stuff does that can't instead be handled by donning a sysadmin hat and managing your agents as separate users?
Like everything else, the packaging and ergonomics matter. Do we need podman or docker when we could just don our sysadmin hats and manage namespaces and cgroups directly instead?
bureado 2 hours ago
The user separation isn’t even necessary, as far as I’ve seen in the projects in https://github.com/bureado/awesome-agent-runtime-security
smashini 4 hours ago
Hey all :)
I've been working an open-source toolkit to stop AI agents from running amok.
You can scan your code (Python, JS, TS) and it will flag any risks and can offer fixes. It runs offline, but you can wire an LLM to do code analysis as well.
You can run it with:
npx @makerchecker/scan
Would love to get any feedback!
smashini 4 hours ago
oopsie doopsie, release pipeline failed fixing now...
smashini 3 hours ago
should be fixed and released now :)
pelagicAustral 3 hours ago
haha! WHAT!? So, we had agents that came with a default setting to request for specific permission to perform an action, then we said "screw it!", we need speed and everybody started coding and releasing agents out in the wild to do whatever they want unchecked... and now we have a product that brings back the safeguards... A few years ago we have abstraction after abstraction coming in the way of blocking actual development (js ecosystem bloat), and now we have layer upon layer for coding with AI...
Zie_Mordecai an hour ago
Really great idea, simple yet effective.
ucsandman 2 hours ago
this is cool I'm working on a similar project called DashClaw. Great work!
christkarani 2 hours ago
great stuff working in a similar project that enforces guardrails at runtime