The Agent Skills Directory

[DYNAMIC_CONTEXT_INJECTION]: The skill uses the ! syntax in SKILL.md to execute a Python script (python3 -c "exec(...)") during the skill loading phase. This is used to automatically check for the presence and version of the yfinance library in the execution environment.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted external data from Yahoo Finance which could be manipulated by third parties to influence agent behavior.
Ingestion points: Headlines and content from the ticker.news API call in SKILL.md.
Boundary markers: Absent. The skill does not use delimiters or instructions to ignore potential commands embedded in news titles.
Capability inventory: The skill demonstrates the ability to execute shell commands via subprocess.check_call and dynamic context injection.
Sanitization: None. The external news data is processed and presented to the agent without filtering or escaping.
[EXTERNAL_DOWNLOADS]: In SKILL.md, the skill includes logic to download and install the yfinance package from the Python Package Index (PyPI) using pip if the library is not already installed.
[COMMAND_EXECUTION]: The skill uses subprocess.check_call to run package management commands and utilizes shell-based commands for environment discovery during the load process.
[DYNAMIC_EXECUTION]: The environment check located in SKILL.md employs the exec() function within a Python one-liner to dynamically evaluate code that checks for module availability.

earnings-recap