The Agent Skills Directory

[COMMAND_EXECUTION]: The skill documents the use of Python's subprocess module to orchestrate the sn and ledger CLI tools. The provided code patterns (e.g., in references/implementation-patterns.md) correctly use list-based arguments for subprocess calls, which mitigates shell injection risks.
[EXTERNAL_DOWNLOADS]: The bot architecture is designed to ingest live trading signals from a remote NATS stream via the sn signals command. This external connection is central to the bot's functionality and is documented as a core architectural component.
[PROMPT_INJECTION]: The skill describes a data ingestion surface for processing external market signals. ● Ingestion points: The bot reads signal data from the sn signals subprocess stdout. ● Boundary markers: Data is processed as structured JSON lines. ● Capability inventory: The bot executes financial transactions and position queries through the ledger CLI. ● Sanitization: The patterns demonstrate parsing input as JSON and validating specific fields, such as numerical bounds checks for stop-loss/take-profit levels and prefix matching for strategy verification.
[NO_CODE]: This skill is primarily documentation and reference-based; it does not include executable script files or binaries within the package, relying instead on the user to implement the described patterns.

trading-bot