at-email-cli

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

• Third-party content exposure detected (high risk: 0.85). Outsider-authored free text can enter the LLM context via mailbox message content (subjects/snippets/bodies/links) returned by at-email inbox/at-email read, which the workflow instructs to treat as untrusted data but still requires reading/relaying it to the agent.

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

• Potentially malicious external URL detected (high risk: 1.00). The skill instructs running a runtime npx command that fetches and executes remote code—"npx --yes @agentteamhq/email@latest" (and references the GitHub release agentteamhq/agentteam-email)—so it relies on externally fetched code executed at runtime.