container-scan-dockle

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

• Third-party content exposure detected (high risk: 0.90). This skill instructs the agent to run Dockle on a user-specified container image and parse Dockle's JSON output (dockle-results.json), which means it ingests untrusted third-party content from container images that may come from public registries or user-provided sources (e.g., Docker Hub), allowing attacker-controlled metadata or file contents to be interpreted by the agent.