Very much WIP, would appreciate any feedback. https://github.com/coproduct-opensource/nucleus

This can be configured via profiles to more more or less restrictive.

Treat this as an example for now, more to come in the future.

Comments URL: https://news.ycombinator.com/item?id=47582980

Points: 2

# Comments: 1

Runtime enforcement means that any side effects are routed through a proxy (nucleus-tool-proxy) that does realtime checks on permissions and gates the behavior.

SPIFFE for MicroVM agents is a compelling idea and I'll update when this is ready.

Nucleus is an OSS experiment that pairs a small, compositional permission model with runtime enforcement: *side effects are only reachable through an enforcing tool proxy*, inside a Firecracker microVM. The envelope is *non-escalating*: it can only tighten or terminate, never silently relax.

What works today:

* MCP tool proxy with *read / write / run* (enforced inside the microVM) * default-deny egress + DNS allowlist + iptables drift detection (fail-closed) on Linux * time + budget caps enforced * hash-chained audit log + HMAC approval tokens (scoped, expiring) for gated ops

What’s missing (being upfront):

* web/search tools exist in the model but aren’t wired to MCP yet * remote append-only audit storage + attestation are still roadmap * early/rough; targeting “safe to run against sensitive codebases,” not “replace your local terminal”

Most of the code was written with Anthropic tools; I’ve been leaning on tests/fuzzing/proptests to keep it honest.

Would love feedback on: (1) dangerous capability combinations beyond the lethal trifecta, (2) what enforcement gaps you’d want closed first, (3) how you’d evaluate this vs gateway-only approaches.

Comments URL: https://news.ycombinator.com/item?id=46855770

Points: 3

# Comments: 3

Comments URL: https://news.ycombinator.com/item?id=12785523

Points: 2

# Comments: 0

I wrote a limited visualizer for this a few years back.

http://bl.ocks.org/bcrisp/6072748

In this the protagonist discovers he is special, exceptional. He devotes time to thinking and researching career goals and world-changing ideas. Dreaming of reification and being lauded as the (Gauss|Galois|Mozart) of his generation, the present may be bleak, but the future is glorious.

Act II: The Crisis

In this, the protagonist discovers that he is challenged for his position. Others, caught up in the race of life, don't look deeply to see his talent, and communion with the muses is precious only as long as it relates to engineering goals. He is human, suffering financial and relational setbacks, feeling the muses have deserted him and he is fully mortal after all.

Act III: The Resolution

In this, the protagonist discovers he is not alone. Others around him, few to be sure, are equally talented and working to achieve their purpose. Perhaps he's at the 99.9th percentile and realizes there are still 6 million contemporaneous peers. Einstein, Crick, and Jobs all went through this before their breakthroughs. The protagonist finds a passion, gives it his heart and soul, and achieves Movement I of his life story.

Yes, a startup would certainly offer the complexity I'm after. I'm not pursuing insurmountable complexity in its own right. I feel we should be doing programming at a vastly higher level of abstraction.

Maybe I ought to shift to Haskell development, or spin higher level of abstraction into a venture of its own. Thanks again.

I want bijective and injective, bipartite and clique, functors and connectedness and small world effects.

I don't want comfort. I want God, I want poetry, I want real danger, I want freedom, I want goodness, I want sin.

Comments URL: https://news.ycombinator.com/item?id=7671485

Points: 2

# Comments: 2