The Agent Skills Directory

[NO_CODE]: The skill consists entirely of instructional Markdown content. It does not contain any scripts (Python, JavaScript, shell), binaries, or configuration files that would execute on the host system.
[DATA_EXFILTRATION]: There are no network-related commands (such as curl or wget) or external URL references. While the skill asks users to provide financial data (account balances, P&L), this information remains within the local agent context and is not transmitted externally.
[PROMPT_INJECTION]: The instructions do not use adversarial patterns such as 'ignore previous instructions' or attempts to bypass agent safety filters. The language is purely instructional for the intended purpose of risk management.
[EXTERNAL_DOWNLOADS]: No external packages, remote scripts, or third-party repositories are referenced or installed during the skill's operation.
[COMMAND_EXECUTION]: The skill does not perform any shell command execution or system-level operations. All calculations and logic are handled through natural language processing of user inputs.

risk-manager