The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill's architecture presents a surface for indirect prompt injection due to the combination of data ingestion and high-impact capabilities.
Ingestion points: The skill is designed to fetch and process untrusted external data, specifically market news and asset metadata through the news and assets toolsets mentioned in SKILL.md.
Boundary markers: There are no explicit instructions or delimiters defined to prevent the agent from following instructions that might be embedded within retrieved market news or ticker descriptions.
Capability inventory: The skill enables the agent to perform sensitive actions including submitting market/limit orders (submit_order), closing positions (close_position), and accessing private account balances.
Sanitization: The instructions do not describe any mechanisms for sanitizing or filtering external content before it is processed by the agent's logic.
[EXTERNAL_DOWNLOADS]: The skill facilitates the retrieval of external software and configuration.
Evidence: It instructs the user to install the alpaca-mcp-server package using the uvx tool and provides instructions for cloning the official Alpaca headquarters repository from GitHub to build Docker images.

alpaca-trading-mcp-server