The Agent Skills Directory

[SAFE]: The skill uses the official cx CLI to interact with Coralogix infrastructure. All operations described, such as asking questions to the AI agent and retrieving artifacts, are standard features of the observability platform.
[COMMAND_EXECUTION]: The skill relies on the cx command-line tool. The instructions provide examples of how to correctly parameterize these commands, including handling long-running investigations by backgrounding processes. There are no attempts to execute unauthorized commands or escalate privileges.
[INDIRECT_PROMPT_INJECTION]: The skill processes data returned from the remote Coralogix Observability Agent (Olly) and fetches artifact content. While these are ingestion points for external data, they represent the core intended functionality of the skill. Standard guardrails for processing remote LLM output apply.
[DATA_EXPOSURE]: The skill interacts with telemetry and observability data (logs, metrics, alerts) which is the primary purpose of a Coralogix tool. It does not attempt to access sensitive local system files like SSH keys or environment secrets.

cx-olly