Experimental system
EvMind is the research programme. EvAGI is the experimental system the programme is built on — the architecture where live learning is demonstrated rather than only measured.
Experimental — research prototype — results are reported, not settled
The two names are often used together because they answer different questions. EvMind asks whether continual learning can be improved — what the hypothesis is, what the baseline is, and what the measurements show. EvAGI is the thing those measurements are taken on.
Keeping them apart matters for reading results honestly. A finding on EvAGI is a statement about one experimental architecture under one evaluation protocol. It is not a finding about continual learning in general, and it is not a claim that catastrophic forgetting has been solved.
EvAGI also sits closer to the engineering side of the ecosystem than EvMind does, because the same harness concepts — routing between specialists, shared tools, persistent memory — appear in EvAgent and EvMem.
The programme this system belongs to, and the index that tracks what has actually been published.