AI Sell Recommendation

Endpoint /v1/market/ai-sell-recommendation — product line ai_assets

Run:

AI Sell Recommendation

PL3 endpoints have no polish layer.They are computed from ta_indicator_scores, not from bithumb_reports, so there is no pre-polish snapshot to compare against. All checks here run against the polished source only.

T0 · Red-line

—

expected but 0 rows

Static / mechanical compliance — any failure means the output is malformed

T1 · Substance

—

expected but 0 rows

LLM-judged correctness + reasoning — hallucination, argumentation, etc.

T2 · Structure

—

expected but 0 rows

Conclusion structure, multi-asset differentiation, vocabulary

T3 · Polish

—

no data

Reader actionability + Korean naturalness + passive-voice ratio

No check results for /v1/market/ai-sell-recommendation in this run