Chanlun Trading System

Problem

Chanlun is often reduced to indicator crossovers, MACD divergence, or breakout buy signals, without level, structure, or invalidation checks. That makes whipsaws hard to avoid. This skill turns analysis into a reviewable pipeline: declare trade_level, confirm_level, and trigger_level, then build structure as containment -> fractal -> stroke -> segment -> center before classifying first, second, or third buy/sell candidates.

It requires fixing the operating level first, then identifying trend type, center, divergence, and trigger level, so a low-level divergence is not promoted into a high-level reversal and indicator state is not mistaken for structure confirmation.

How it works

• Gated reasoning: level_gate, structure_gate, comparison_gate, and risk_gate force a signal to downgrade to observe or proxy_research when a step is missing.
• Mode layering: separates strict_chanlun, structure_proxy, and proxy_research, so approximate inputs are not mistaken for original definitions.
• Signal discipline: each entry must state evidence, trigger, invalidation, and next observation point; MACD, RSI, and volume adjust confidence or sizing only.
• Backtest auditability: requires definition_mode, structure_completeness, and approximation_loss, then checks common false positives.

Boundaries

Best for technical research, rule organization, and backtest proxy design; not for direct trade instructions. When data is sparse or structure is unclear, the correct action is to wait rather than force a trade; when charts are missing, output assumptions only, not a fake visual reading.