1panel-api

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.80). The skill's API docs show endpoints that fetch and ingest arbitrary external content—e.g., cronjobs supports scriptMode=scriptUrl and a url field (docs/SKILL-cronjobs.md), apps support syncing a remote app store (docs/SKILL-apps.md), and container/image pull and repo downloadUrl allow pulling images from public registries (docs/SKILL-containers.md)—which the agent would read/act on and could therefore enable indirect prompt injection.

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 0.80). The skill exposes 1Panel APIs that include user management, hosts/SSH, firewall, service and file management — functionality that can modify system state (including creating users or changing SSH/service configs) and thus could be used to compromise the machine the agent runs on.