The Agent Skills Directory

[SAFE]: The skill's operations are confined to its designated workspace and utilize official platform tools for trading functionality. No malicious behaviors such as prompt injection, data exfiltration, or unauthorized persistence were observed.
[COMMAND_EXECUTION]: Script 'scripts/phoenix_config.py' uses subprocess.run to execute the platform's 'mcporter' CLI tool. This is used exclusively for fetching market data and account state, following safe implementation practices (using argument lists instead of raw shell strings).
[DATA_EXFILTRATION]: All external data retrieval is performed via authenticated platform tools to known providers. No sensitive information, such as environment variables or credentials, is accessed or transmitted to unauthorized third parties.

phoenix-strategy