The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a local Python script (trailing_stop.py) using the uv tool to interact with the Interactive Brokers API on local ports (7496/7497). This is the core functionality intended for the skill.\n- [EXTERNAL_DOWNLOADS]: The skill lists trading-skills as a dependency, which is an internal library associated with the vendor's toolset.\n- [PROMPT_INJECTION]: The skill ingests position and order data from the Interactive Brokers API (Ingestion) to generate formatted markdown reports and manage orders (Capabilities). This creates a surface for indirect prompt injection via asset names or notes, though the risk is mitigated by the skill's specific purpose and the requirement for explicit execution flags.\n
Ingestion points: Portfolio and order data from Interactive Brokers API.\n
Boundary markers: Structured JSON output formatted into Markdown reports.\n
Capability inventory: Order placement and cancellation via the broker API.\n
Sanitization: Relies on the trading_skills library and agent-level reporting logic.\n- [SAFE]: The skill implements a 'dry-run' safety mechanism by default; state-changing operations like placing or canceling orders are only performed if the user explicitly provides the --execute flag.\n- [SAFE]: Configuration of sensitive connection details, such as the IB_PORT, is handled via environment variables or .env files, which follows security best practices for local secret management.

ib-trailing-stop