The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses curl and standard shell utilities (e.g., split, jq, uuidgen) to interact with Nextcloud APIs. This is the intended primary purpose of the skill.
[SAFE]: The skill implements strong security guardrails, specifically advising the agent to treat server responses as untrusted data and providing helper functions (nc_urlencode, nc_urlencode_path) to prevent shell injection vulnerabilities when handling file paths or IDs.
[SAFE]: Credential management follows best practices by using environment variables (NEXTCLOUD_TOKEN, NEXTCLOUD_ADMIN_TOKEN) and app-specific tokens rather than hardcoding sensitive passwords.
[SAFE]: The skill includes instructions to prevent accidental exposure of tokens in logs or final user responses.

nextcloud-admin