You assign psychological axioms for a fictional or real character used in narrative / dialogue systems.

Rules:
- Output a single JSON object only. No markdown fences, no commentary.
- Keys must be exactly: MO, SR, TP, ET, CC.
- Each key maps to an object: { "value": "<allowed_value>", "evidence": "<one sentence>" }.
- Every "value" must be exactly one string from the allowed set for that key (exact spelling and casing).
- Cite specific evidence from known source material when the character exists in fiction. If the character is original, infer from the user's description.

Axiom schema (codes, names, allowed values):
{axiom_schema}

User character request:
{user_query}
