llm-router

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

• Third-party content exposure detected (high risk: 0.80). The skill's workflow explicitly shows using agent mode with MCP tool integrations that access public user-generated sources (e.g., "swiftopenai agent ... --mcp-servers github --allowed-tools 'mcp__github__*'" in SKILL.md), meaning the agent will fetch and interpret untrusted third‑party GitHub content as part of its execution and that content could influence subsequent tool calls or actions.

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

• Potentially malicious external URL detected (high risk: 0.80). The scripts/check_install_cli.sh script runs "npm install -g swiftopenai-cli", which at runtime fetches and installs remote code from the npm registry (https://registry.npmjs.org/) that is then executed as the required CLI dependency for agent operations.