mrajagopalan

2 ms·

1. 1.▲
   Authenticated Workflows: A Systems Approach to Deterministic Agentic Controls(arxiv.org)

   3 points by mrajagopalan 8mo ago|1 comments
2. 2.▲
   by mrajagopalan 8mo ago

   We reduce agentic systems to byzantine distributed systems and protect 4 boundaries — prompts, tools, data and context -- via a trust layer that enforces intent (operations satisfy organizational policies) and integrity (operations are cryp
3. 3.▲
   by mrajagopalan 8mo ago

   OpenClaw highlighted the core problem with securing AI agents: you can't enumerate every malicious prompt. Pattern matching catches "rm -rf" but misses "please remove all files." Distributed systems solved this deca
4. 4.▲
   Show HN: Applying OAuth/RBAC patterns to secure AI agents(macawsecurity.ai)

   2 points by mrajagopalan 8mo ago|1 comments
5. 5.▲
   by mrajagopalan 1y ago

   A bit late to this discussion — but we've been looking at this problem for a while and have implemented a cryptographic approach I wrote about here: https://news.ycombinator.com/item?id=45244297_ID TL;DR: We treat AI c
6. 6.▲
   Setting Boundaries: Getting Zero-Trust Tool Calling Right for Agentic AI(macawsecurity.com)

   2 points by mrajagopalan 1y ago|1 comments
7. 7.▲
   by mrajagopalan 1y ago

   Following the MCP security discussion - https://news.ycombinator.com/item?id=45199713 I've written up our approach to securing AI tool calling through cryptographic enforcement rather than prompt engineering. The core p