claude-channels

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

• Third-party content exposure detected (high risk: 0.90). The skill's Channels workflow (see references/custom-channels.md, references/telegram.md, references/discord.md and security-and-permissions.md) explicitly ingests untrusted third‑party content — Telegram/Discord messages, CI/webhook payloads, and external log/artifact URLs — and instructs Claude to read and act on that content (fetch logs, triage alerts, and potentially run/automate commands), which could enable indirect prompt injection.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill's prerequisite installs Bun using a runtime command that downloads and executes remote code (curl -fsSL https://bun.sh/install | bash), which is a required dependency and directly runs external code.