Home / Glossary / EVA — Epistemic Validator Agent

EVA — Epistemic Validator Agent

An epistemological validation agent that evaluates intellectual rigour — separating hypothesis, model and evidence, and flagging wishful thinking before any publication.

An epistemological validation agent. It evaluates the intellectual rigour of frameworks, theses, arguments and conceptual content — explicitly separating what is hypothesis, model or evidence. It identifies validity boundaries, flags wishful thinking, and verifies whether conclusions are supported by their premises. It never presents as "proven" what is merely "plausible."

Context & Strategy

How it was born

It was born as step 7 of the Thought Leadership Pipeline — epistemological validation as a mandatory component before any publication or product launch based on intellectual claims. It was formalised as a specialist agent in February 2026, in the context of developing the agent system for thought leadership production.

What it does in practice

Activated before any whitepaper, framework or thesis publication. It is the complementary pair to the anti-wishful-thinking system: while the latter audits already-produced frameworks, EVA validates whether the claims being made are epistemically sustainable. It also complements RIA — when analysing something external, EVA assesses whether the analysis produced is itself rigorous.