The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill handles sensitive KTX/Korail login credentials but implements robust security practices. It prioritizes environment variables and secret vaults (like 1Password or Bitwarden) and provides a fallback to a local configuration file (~/.config/k-skill/secrets.env) with appropriately restricted permissions (0600).
[EXTERNAL_DOWNLOADS]: The skill installs the korail2 and pycryptodome packages from the official Python Package Index (PyPI) if they are missing. These are well-known and standard libraries for interacting with Korail services and performing cryptographic operations.
[COMMAND_EXECUTION]: The skill executes a local helper script (scripts/ktx_booking.py) using shell commands to perform train searches, reservations, and cancellations. This is a controlled execution of the skill's own components.
[PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it processes data retrieved from the Korail API.
Ingestion points: Data returned by the search and reservations commands of the ktx_booking.py script.
Boundary markers: No specific delimiters or boundary instructions are mentioned for the processing of train schedule or reservation data.
Capability inventory: The skill can execute subprocesses (the helper script) and read local configuration files for credentials.
Sanitization: No explicit sanitization or validation logic for the external API responses is described in the workflow.
[DATA_EXFILTRATION]: No malicious data exfiltration patterns were detected. Network activity is limited to the expected interaction with Korail infrastructure via the documented libraries.

ktx-booking