ralph-install

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

• Third-party content exposure detected (high risk: 1.00). The skill's workflow explicitly instructs fetching/cloning the public GitHub repo at https://github.com/tolulawson/ralph-harness and to read its INSTALLATION.md and src/ manifests as authoritative install instructions, which means the agent will ingest and act upon untrusted third-party content from a public site.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly fetches/clones https://github.com/tolulawson/ralph-harness at runtime and requires reading/applying INSTALLATION.md and the canonical AGENTS.md loader snippet from that repo, which directly control agent prompts/installation behavior.