The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes ae-cli to perform project discovery, dashboard listing, and activity analysis. These commands are standard administrative operations within the supported environment and are not used for malicious purposes.\n- [SAFE]: The skill incorporates a mandatory PROJECT_ID_GATE to ensure that operations only proceed after the target and audit projects have been correctly identified and verified against the environment's actual project list.\n- [SAFE]: Explicit safety instructions require the agent to obtain human confirmation at two separate stages: once after identifying candidates and again before the final execution of any write operations (freezing).\n- [SAFE]: The skill explicitly restricts the agent to using only the ae-cli tool and forbids any external HTTP endpoints, limiting the execution environment to trusted internal tools.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes data such as dashboard and project names. This surface is protected by instructions that mandate using only API-retrieved data and requiring user confirmation of the specific entities before action is taken.

ae-freeze-inactive-dashboards