agent-ready-mcp

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

• Third-party content exposure detected (medium risk: 0.30). SKILL.md Step 4/Step 5 describes using the scan_site MCP tool with a user-provided url, where the server returns scraped text from that target site (e.g., headings, llms.txt/AGENTS.md bodies) that can include outsider-authored prompt-injection-like instructions.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill instructs running remote code via npx ("npx -y agent-ready-mcp@latest") which fetches/executes the npm package at runtime and also offers a runtime remote MCP endpoint (https://agent-ready.dev/api/v1/mcp) that would provide prompts/tools from the remote host, so these external endpoints control agent behavior or execute code.