The Agent Skills Directory

[SAFE]: The skill provides standard MQL5 code patterns for algorithmic trading filters. No patterns of obfuscation, exfiltration, unauthorized command execution, or remote code downloads from unknown sources were identified. All referenced documentation points to internal project resources.
[INDIRECT_PROMPT_INJECTION]: The skill defines an architecture where external market data feeds directly into logic that triggers automated trading actions.
- Ingestion points: Market data queries using iATR, SymbolInfoInteger, GetADX, GetPlusDI, and GetMinusDI within SKILL.md.
- Boundary markers: None; inputs are processed as numerical values for conditional logic gates.
- Capability inventory: Order execution functionality through g_TradeMgr.OpenTrade and trade parameter calculation via g_RiskMgr.CalculateLotSize.
- Sanitization: The logic utilizes standard MQL5 numerical data types and built-in terminal indicator functions which inherent platform-level constraints.

market-regime