I'm building tenbri, an open source record of agents decisions.
The idea came to me after the HF attacks, for anyone building agents going through logs and traces after this kind of attacks is a nightmare and can expose internals for auditors.
This is still very early and I'm working on a proxy mode that alert also in case of silent running agents to also notify when agents decide to skip the logging of decisions.
Looking for honest feedback and all the possible help :)
Hi HN,
I'm building tenbri, an open source record of agents decisions.
The idea came to me after the HF attacks, for anyone building agents going through logs and traces after this kind of attacks is a nightmare and can expose internals for auditors.
This is still very early and I'm working on a proxy mode that alert also in case of silent running agents to also notify when agents decide to skip the logging of decisions.
Looking for honest feedback and all the possible help :)