The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes superset CLI commands (such as status, auth whoami, and hosts list) to gather diagnostic information about host services and user sessions. These are standard operations for troubleshooting the tool.
[EXTERNAL_DOWNLOADS]: The skill utilizes the superset update command to synchronize versioning between the CLI and the desktop application, representing a typical update mechanism for vendor software.
[DATA_EXFILTRATION]: The skill accesses local application log files at ~/Library/Logs/Superset/main.log for error identification. The instructions explicitly mitigate potential data exposure by requiring the agent to obtain user consent before sharing log excerpts in external feedback reports.
[INDIRECT_PROMPT_INJECTION]: The skill evaluates user-provided symptoms to determine appropriate fixes, creating a potential attack surface.
Ingestion points: Symptom descriptions provided as arguments in the skill invocation.
Boundary markers: No explicit delimiters are used to isolate user input from the skill's instructions.
Capability inventory: Access to system-level CLI commands and log files.
Sanitization: No evidence of input validation or filtering is provided for the user-supplied symptom descriptions.

superset-doctor