Oxlamarr

6 ms·

1. 1.▲
   Show HN: Open-source control engine for agents(attestorportal.com)

   1 points by Oxlamarr 2mo ago|0 comments
2. 2.▲
   Show HN: Zero Trust Boundary for Agents(github.com)

   4 points by Oxlamarr 2mo ago|0 comments
3. 3.▲
   A TypeScript prototype admission layer for agents(github.com)

   2 points by Oxlamarr 3mo ago|0 comments
4. 4.▲
   TypeScript control plane for AI driven operations(github.com)

   4 points by Oxlamarr 3mo ago|0 comments
5. 5.▲
   by Oxlamarr 5mo ago

   The harness point is probably the most important part here. With terminal agents, it often feels like the benchmark is measuring the whole loop: model, prompt, tool interface, retry policy, timeout handling, and recovery from bad shell comm
6. 6.▲
   by Oxlamarr 5mo ago

   Very cool. Is the bottleneck under load mostly SQLite write throughput, or the WAL notification layer?
7. 7.▲
   by Oxlamarr 5mo ago

   The speed of progress here is wild. It feels like the hard part is shifting from having access to a strong model to actually building trustworthy systems around it.
8. 8.▲
   by Oxlamarr 5mo ago

   This is exactly why we can't just wrap APIs around LLMs and assume it's secure. The execution layer needs to be completely decoupled from the generation layer. When your proxy or agent framework inevitably gets compromised (like t
9. 9.▲
   by Oxlamarr 5mo ago

   Love this architectural approach. Using probabilistic models to verify other probabilistic models is just turtles all the way down, so anchoring the agent to deterministic AST evidence is exactly the right move. I've been working on th