Looking forward to 0.1.0 release :)","title":null,"type":"comment","url":null},{"author":"xnx","children":[{"author":"billfor","children":[],"created_at":"2026-09-15T17:37:40.000Z","created_at_i":1789493860,"id":49715935,"options":[],"parent_id":49715422,"points":null,"story_id":49713261,"text":"Yeah they could have picked a better name. OpenShell/ClassicShell have been around for a long time (before them).","title":null,"type":"comment","url":null}],"created_at":"2026-09-15T16:59:21.000Z","created_at_i":1789491561,"id":49715422,"options":[],"parent_id":49713261,"points":null,"story_id":49713261,"text":"For a moment I thought agentic AI had really spread to every project, but this is not related to the open-source Windows Start menu replacement: https://github.com/Open-Shell/Open-Shell-Menu","title":null,"type":"comment","url":null},{"author":"zaphar","children":[{"author":"recursivecaveat","children":[],"created_at":"2026-09-15T21:13:33.000Z","created_at_i":1789506813,"id":49718989,"options":[],"parent_id":49716274,"points":null,"story_id":49713261,"text":"I keep seeing the same contradiction: people consider certain activities like "write code that me or my customers run" or "read my email", to be table stakes abilities for an agent. At the same time, they expect some kind of sandbox that prevents even a malicious agent from doing damage. You already gave it the most broad and powerful possible capabilities at the beginning though, there's not even a barn left to close the door on.","title":null,"type":"comment","url":null}],"created_at":"2026-09-15T17:58:05.000Z","created_at_i":1789495085,"id":49716274,"options":[],"parent_id":49713261,"points":null,"story_id":49713261,"text":"I don't really feel like permissions is the hard bit here. The hard bit is that for an agent to do any kind of useful work it needs access to a lot of stuff. Further, the stuff it needs access to probably has vulnerabilities the agent can do. We have not exactly designed our environments for a use case like this. I've gone through the sandboxing exercise for an agent and by the time I had given it enough permissions to do anything useful my sandbox looked like swiss cheese.","title":null,"type":"comment","url":null},{"author":"philipwhiuk","children":[],"created_at":"2026-09-15T18:00:10.000Z","created_at_i":1789495210,"id":49716301,"options":[],"parent_id":49713261,"points":null,"story_id":49713261,"text":"> Back in the 2016 timeframe, members of our team were working at AWS and faced a similar challenge. Given all of the awesome complexity of AWS IAM policies, AWS S3 storage policies, historical version support- can we definitively say whether an object in S3 is accessible to the public internet or not?
To be honest AWS accounts are a complete mess and the AWS dashboard does not gracefully handle reduced permissions so this is not an inspiring start.
The policy framework is everything that explains why formal methods has gotten very little traction in day-to-day software engineering, it is extremely difficult to parse (maybe this is a Z3 problem)","title":null,"type":"comment","url":null},{"author":"wek","children":[],"created_at":"2026-09-15T19:28:09.000Z","created_at_i":1789500489,"id":49717619,"options":[],"parent_id":49713261,"points":null,"story_id":49713261,"text":"Thank you for this. This seems like a promising approach to agent permissioning. What are the performance implications?","title":null,"type":"comment","url":null}],"created_at":"2026-09-15T14:40:05.000Z","created_at_i":1789483205,"id":49713261,"options":[],"parent_id":null,"points":40,"story_id":49713261,"text":null,"title":"What we have learned at OpenShell applying formal methods to control AI agents","type":"story","url":"https://nvidia.github.io/OpenShell-Research/dev-notes/posts/2026-09-10-learning-formal-methods-agent-policy-prover/"}