pp-grubhub

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

• Third-party content exposure detected (high risk: 0.70). SKILL.md describes a runtime workflow where the CLI “browses Grubhub’s marketplace from the command line” and scrapes Grubhub to produce JSON; any scraped text would be outsider-authored web content that can be injected into the agent’s LLM context via the tool’s stdout when using --agent.

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

• Potentially malicious external URL detected (high risk: 0.85). The skill includes go install commands that fetch and build remote code at runtime from github.com/mvanhorn/printing-press-library/library/commerce/grubhub/cmd/grubhub-pp-cli@latest and github.com/mvanhorn/printing-press-library/library/commerce/grubhub/cmd/grubhub-pp-mcp@latest, which execute remote code as required installation steps before the skill can run.