niyikiza

5 ms·

1.
by niyikiza 17d ago
To be fair hardly anybody really knows what it means to be a regular member either.
2.
Per-request authorization for Temporal Nexus(tenuo.ai)
1 points by niyikiza 1mo ago|0 comments
3.
by niyikiza 1mo ago
This. Especially when you have agents calling other agents. Just published an article about that yesterday: https://niyikiza.com/posts/agents-to-agents/
4.
Ordinary agent tool calls can create shadow delegation(niyikiza.com)
2 points by niyikiza 1mo ago|0 comments
5.
by niyikiza 3mo ago
We built an open source governance tool called Tenuo that allows you to define exactly what an agent can do for a given tasks. Works for local filesystem, network, mcp servers, etc.
6.
by niyikiza 3mo ago
I'm working on a Claude Code governance tool that allows to define deterministic policies for tool call that can be enforced across a fleet and will be in effect even when individual users run with --dangerously-skip-permissions https
7.
by niyikiza 3mo ago
I guess they should include tuition cost as well.
8.
by niyikiza 3mo ago
We've been using Tenuo which for task-scoped authorization. Its integration for Claude Code: https://github.com/tenuo-ai/claude-governance
9.
by niyikiza 3mo ago
Because procurement is hard. Changing vendors is a big undertaking for big companies. They are certainly not going to be switching vendors every time there is an incident
10.
by niyikiza 3mo ago
Means you can basically host your own AS
11.
by niyikiza 3mo ago
there are some emerging mechanisms for offline verification that don't require AS in the OAuth WG. (I'm working on one of them)
12.
by niyikiza 4mo ago
What would the benefit be? A mega agent that does everything? There are some well documented advantages of decomposition...that's why the industry favours microservices over monoloths.
13.
by niyikiza 4mo ago
I agree with the coarse permissions point, and I wouldn't bank on those services adding finer-grained scopes. The idea in my draft is to do the attenuation and verification before the call reaches the service, enforced at the boundary,
14.
by niyikiza 4mo ago
There's some active discussions on task level authz and multi-hop delegation in the OAuth WG right now. WorkOS wrote a good overview of the open drafts [1]. (Disclosure: one of them is mine.) [1] https://workos.com/blog
15.
Zero-Touch OAuth for MCP(blog.modelcontextprotocol.io)
278 points by niyikiza 4mo ago|103 comments
16.
by niyikiza 4mo ago
Some clever workarounds wrt process management & POSIX compatibility
17.
by niyikiza 5mo ago
Agree with the meta point. I worked in Korea and Japan and loved the culture but when I moved to the west I was surprised to see how people over here fantasize about their (imo inefficient) corporate cultures. This particular article was de
18.
by niyikiza 5mo ago
My understanding is that when it's something that requires user action they'd directly send comms to customers.
19.
by niyikiza 5mo ago
Probably the best option after sending a mass email when customers need to take action. The status page is for reliability issues impacting end users & the blog is for in-depth analysis.
20.
by niyikiza 5mo ago
Building tenuo.ai ( https://github.com/tenuo-ai/tenuo ): task-scoped authorization for AI agents. Rust implementation of capabilities + cryptographic offline verification.
21.
by niyikiza 5mo ago
My analogy[1] has been that we need a valet key: capped speed, geofenced, short ttl, can't open trunk/glovebox, etc. That way we don't have to say pretty please to the valet and hope that they won't get ideas. [1] https
22.
by niyikiza 5mo ago
I have to agree here...of all things that went wrong here, I don't think the API surface is to blame. You need to have deterministic control & escalation mechanism on your agents whether they are calling an API or any other tool
23.
by niyikiza 5mo ago
I SAY AYE.
24.
by niyikiza 5mo ago
Yeah, people calibrate trust to the median behaviour of the model and get burned by the tail. What makes it harder is that even people who do see the holes often respond with better prompts and more elaborate context. Same trust-the-model m
25.
by niyikiza 5mo ago
Two things get called "hooks" here. Exit code 2 + stderr is a real control. JSON in stdout degrades to a string in the model's tool-result context, where the model is correctly trained to resist instructions because that&#x27
26.
by niyikiza 5mo ago
>>harnesses should have more assertive layers of control and constraint Been saying this for a while and mostly getting blank stares. In-context "controls" as the primary safety mechanism is going to be a bitter lesson for o
27.
by niyikiza 5mo ago
Reminds me of the riddle[1][2] from Game of Thrones / A Clash of Kings: Lord Varys: Three great men sit in a room: a king, a priest, and a rich man. Between them stands a common sellsword. Each great man bids the sellsword kill the oth
28.
by niyikiza 6mo ago
Speaking of fantansies...another approach would be holder binding: DPoP (RFC 9449) has been stable for a couple of years, AWS SigV4 does it too. The key holder proves control at call time, so a captured token without the key is useless.
29.
by niyikiza 6mo ago
I took that course too and ruined my life...by making me think writing a compiler could be fun. The course itself was worth the money I paid for the program.
30.
by niyikiza 6mo ago
Was just at [Un]prompted conference where this was a live debate. The conversation is shifting but not fast enough. I've been screaming about this for a while: we can't win the prompt war, we need to move the enforcement out of th

More ›