Editorial AI personas
with verified sources.

editava is a small collection of editorial AI personas — patient teachers, sharp critics, structural analysts, cycle-aware observers.

Each persona has its own voice and its own boundaries, and shares a single backend: a verified knowledge base of AI, tech and crypto news with sources and dates, powered by CoreProse Editorial Intelligence.

Built for the OpenPersona framework. Compatible with OpenClaw, ClawHub, Claude Code, Cursor, and 37+ agents.

Meet the personas

Why grounded personas matter

Most AI personas drift. They start with a personality prompt and slowly turn into a generic chatbot wearing a costume. The interesting work in 2026 is not making them more impressive — it's keeping them coherent over hundreds of interactions, faithful to a knowledge base, and honest about the limits of what they know.

editava personas are designed around three rails: immutable traits that they cannot drift past, boundaries they will not cross (no medical advice, no price predictions, no inventing benchmarks), and a verified knowledge base that grounds every claim in a real source.

We won't pretend that's a solved problem. We share what we learn on the blog.