detecting-container-runtime-threats-with-falco

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

• Third-party content exposure detected (medium risk: 0.30). In scripts/agent.py, the runtime --triage <alerts.json> mode ingests outsider-provided free text from a user-supplied JSON file path (json.loads(line) per line) to summarize Falco alert fields.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.80). The skill contains runtime install commands that fetch and install/execute remote packages (helm repo add https://falcosecurity.github.io/charts, curl -fsSL https://falco.org/repo/falcosecurity-packages.asc, and the apt source https://download.falco.org/packages/deb), so external content is retrieved and executed as a required dependency.

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

• Attempt to modify system services in skill instructions detected (high risk: 1.00). The skill explicitly instructs running sudo commands, modifying system files (e.g., /usr/share/keyrings, /etc/apt/sources.list.d, /etc/falco/), and enabling system services with systemctl, which require privileged changes to the host and can alter machine state.