I'm building the agentic runtime and agents running on this platform. The keystone of agentic work is "allowlists vs denylists" choice. On my platform I use allowlists for everything (tools, MCPs, skills, memories etc.). Agents work in a sandbox and have only allowed resources they need for functioning. Nothing more.
The LLM is a probabilistic by nature, it can try to access or delete the resource without a reason, it may follow the prompt injection, it may leak the secret or private data. So all this should be controlled by the agentic runtime.
In my system I use specialized agents as team players - each agent does it's part of work using available resources. This way you can keep all needed data in the model's context window without compaction and optimize resources control.
Ask more specific question if you need a detailed answer.
denn-gubsky
I don't want to be mean, but you sound confused.
Are you building agents? or building apps? Or building an API platform that has everything from databases and auth and coffee making capabilities?
Sorry for meandering, but this is a common issue, way more people build builders and sell shovels than there is a demand for, and it's usually a way to escape actually building an application.
Try building something concrete, and asking concrete questions about that.
TZubiri
I want to build an SEO agent with access to SERP and keyword volume data
comments (3)
denn-gubsky
Are you building agents? or building apps? Or building an API platform that has everything from databases and auth and coffee making capabilities?
Sorry for meandering, but this is a common issue, way more people build builders and sell shovels than there is a demand for, and it's usually a way to escape actually building an application.
Try building something concrete, and asking concrete questions about that.
TZubiri
kdqed