Brian Hall

Co-Founder @Faramesh_Labs | Governance-as-Code for AI agents

[Brian Hall @brichall_]

An AI agent with real access and no enforcement is essentially a very confident intern with root.. here's how to put guardrails that ACTUALLY work on a LangGraph agent.
https://t.co/wnc9gZifzu

[Brian Hall @brichall_]

Yep. But once it's touching real systems, seeing what it did isn't enough, you want to gate what it can do too

[Brian Hall @brichall_]

@LangChain @MukilLoganathan Good for blast radius. The gap is still per-action, it can have legit access and still run something it shouldn't

[Brian Hall @brichall_]

Looking to meet more people building around AI agents.

Agent infra, MCP, security, devtools, deployment, enterprise workflows, or anything adjacent.

Especially interested in people trying to get agents into prod without giving them insane levels of access.

If you’re somewhere in that mess too, I’d love to know what you’re building :)

[Brian Hall @brichall_]

@Garcia_Capital Yep. Let the agent suggest the move, but don’t give it the keys to the wallet

[Brian Hall @brichall_]

@boubacarbarry Spot on. People talk about better models like they’ll magically fix agent security, but it cuts both ways. Smarter agents with access to more powerful tools also make prompt injections more creative, less obvious, and harder to catch.

[Brian Hall @brichall_]

Watching the spurs blow a 29 point finals lead and realizing this was preventable with like four lines of policy

agent "deaaron-fox" {
rules {
deny layup if clock < 24 and lead > 0
permit hold_ball
defer anything_involving_outrunning("OG Anunoby")
}
}

Game 5 Saturday. Somebody apply the governance file

[Brian Hall @brichall_]

@AbhiCodes15 My cofounder and i talk about this all the time

Building the thing is somehow the easy part

Getting the right people to give a shit at the right time is the key

[Brian Hall @brichall_]

What are the best lowkey startup accelerators people don’t talk about enough?

Not YC, a16z, Techstars, etc.

We’re building in AI governance / infra if that changes your answer

[Brian Hall @brichall_]

@neuralunlock i went to a panel hosted by a pre-seed/seed VC firm and one of them said “we don’t need startups, startups need us” with a straight face

i wish he was kidding..

[Brian Hall @brichall_]

Why is my whole feed "ai loops" right now

Pretty awesome idea until the loop has tools, credentials, and 9 hours alone with your cloud bill

[Brian Hall @brichall_]

@LangChain Observability is pretty pointless when it’s operating alone, it basically just rubs in how you let the system run poorly.

Being able to enforce policy and stop those loops before they happen is the real unlock... with observability baked in ofc

[Brian Hall @brichall_]

AI agent demo: wow this is the future
AI agent asking for prod access: absolutely not

[Brian Hall @brichall_]

@garrytan The agent era is just everyone rediscovering why permissions existed in the first place

[Brian Hall @brichall_]

https://t.co/nbGSh22SmW

Article is here if you want to read it. It's always cool when people in the space find Faramesh on their own

[Brian Hall @brichall_]

Nice to see @faramesh_labs show up here.

Hot take that probably isn't that hot: agentic governance gets won by deterministic enforcement, not probabilistic judging.

Once you care about non-bypassability, runtime enforcement before the action is what matters.

[Brian Hall @brichall_]

@larsencc @rohit_jsfreaky The agent shouldn’t get some giant permission blob and then everyone hopes for the best

Every action should get checked when it happens... this agent, this context, this tool call. allowed or not.

That’s the layer (very broadly) IMO

[Brian Hall @brichall_]

@dessaigne Yup. Now the fun part is making it safe for those agents to actually use all those tools SAFELY

[Brian Hall @brichall_]

@Yuchenj_UW NYC Tech Week has been testing this

So many good rooms to be in, but at some point you have to pick the ones that actually matter and not feel bad about skipping the rest