agent-tools

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

• Third-party content exposure detected (high risk: 0.80). The skill documentation (SKILL.md and references/app-discovery.md) explicitly shows running web-search and extraction apps (e.g., "belt app run tavily/search-assistant" and listings for Tavily/Exa Search/Exa Extract) which ingest public web content that the agent is expected to read and could influence subsequent actions.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill instructs users to run a remote installer with "curl -fsSL https://cli.inference.sh | sh", which fetches and executes remote code from https://cli.inference.sh and is a required runtime dependency for the CLI-based workflow.