The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/schedule.py executes shell commands via subprocess.run to interact with system utilities such as crontab and schtasks. These operations are used to manage automated alerting schedules. Additionally, scripts/notify.py uses subprocess.run to trigger desktop notifications using the notify-send binary on Linux systems.
[PERSISTENCE]: The scripts/schedule.py module allows for the creation of persistent background tasks. It can configure a cron job on Unix-based systems or a Scheduled Task on Windows to execute the alert evaluation script periodically (every 5 minutes on weekdays) without requiring an active agent session.
[EXTERNAL_DOWNLOADS]: The scripts/quotes.py file fetches live financial data from various external APIs and services, including Yahoo Finance, Alpaca Markets, Finnhub, Alpha Vantage, and Nasdaq. These are well-known financial data providers used for the skill's primary functionality.
[DATA_EXFILTRATION]: Through scripts/notify.py, the skill transmits alert data to external endpoints. It supports sending ticker symbols, price information, and alert messages to user-defined URLs via Slack, ntfy.sh, or custom webhooks. This behavior is documented as a core feature of the alerting system.
[INDIRECT_PROMPT_INJECTION]: The skill ingests structured and unstructured data from external financial APIs. This data is subsequently processed and presented to the agent. While the data is primarily financial metrics, the ingestion of ticker names and earnings summaries from third-party sources represents a surface for potential indirect prompt injection, although the risk is limited by the structured nature of the data processing pipeline.

price-alerts