Mohamed Azahrioui
backend developer, the hague
I’m a backend developer in The Hague, two years into a CS degree at Leiden and already writing production code for real clients. What I care about is correctness: code that stops when something is wrong instead of guessing, and that keeps a record you can check.
guardrails
Three times now I’ve built roughly the same thing. Each one sits in front of a spot where an AI system would otherwise guess, and puts a fixed rule in charge: the rules decide, the model just explains the decision, and you get output you can check instead of trust. I didn’t plan to specialise in this. It just kept being the problem in front of me.
RETRIEVED / FAILED / UNKNOWN
This is the one I put on PyPI. It sits where an agent fetches a web page and checks whether the page was actually read, or whether the request just came back with a 200. A 200 only means bytes arrived, not that you got the page. If it can't confirm a real read, it stops the agent instead of letting it answer anyway. No model, no API keys. One test in the suite deliberately tries to slip a fake read past it, and fails.
REACHABLE / UNKNOWN / NOT-AFFECTED
A scanner tells you a vulnerable function exists somewhere in your dependencies. It doesn't tell you whether your own code can ever reach it. On the codebases I tried, roughly four in five “critical” findings couldn't be reached at all. ReachGate walks the code graph to check whether a path exists. If it runs out of budget before it can be sure, it returns UNKNOWN rather than marking something safe it never proved. Every result exports as signed proof you can re-run offline.
ALLOW / APPROVAL_REQUIRED / BLOCK
Before an agent does something costly, like issuing a refund, TrustGate looks at the data behind the decision: is it fresh, does it still match the contract it was written against, is the source even connected? An agent can be allowed to act and still be acting on bad data. It returns allow, needs-approval, or block, with the evidence attached.
other things I’ve built
writing
all posts →