The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses standard CLI tools (curl, jq, journalctl) to interact with the healthchecks.io service. These operations are within the expected scope for managing and monitoring services.
[CREDENTIALS_UNSAFE]: The skill correctly instructs the user to provide credentials via environment variables ($HC_API_KEY, $HC_PING_KEY) instead of hardcoding sensitive information, which is a recommended security practice.
[DATA_EXPOSURE]: The skill includes an example of sending service logs to a ping URL using journalctl. While this involves transmitting local system data to an external endpoint, it is a documented and intended feature of the healthchecks.io service for diagnostic purposes.
[INDIRECT_PROMPT_INJECTION]: The skill processes JSON responses from an external API. The use of jq for parsing helps mitigate risks associated with untrusted data ingestion, and the impact is limited to the agent's internal processing of monitor statuses.

healthchecks-io