Security audit
Security checks for vulnerabilities and agentic risk
The skill does what it says by sharing a live website preview, but its install and tunnel instructions expose users to avoidable high-impact risk.
Review before installing. Prefer the pip or uv package path in an isolated environment, avoid the curl-to-bash and PowerShell iex installers, and only tunnel non-sensitive local previews. Do not expose admin/debug routes, secrets, private repositories, or production data through the public preview link, and stop the server and tunnel when finished.
SKILL.md:50Unverified Remote Installer Scripts Are Executed Directly
SKILL.md:14Unpinned Third-Party Packages May Execute Unreviewed Code
curl -fsSL https://aitun.cc/install.sh | bash fetches and immediately executes remote code with no opportunity for review or integrity verification. If the remote server or delivery path is compromised, this becomes arbitrary command execution on the host running the skill.
Or install via one-line script (Linux/macOS):
curl -fsSL https://aitun.cc/install.sh | bash
Windows (PowerShell):
The | bash construct is a classic dangerous chaining pattern because it turns downloaded network content directly into executable shell input. In an agent setting, this is more dangerous than a normal documentation snippet because it may be copied or executed automatically without human inspection.
Or install via one-line script (Linux/macOS):
curl -fsSL https://aitun.cc/install.sh | bash
Windows (PowerShell):
Repeating the remote-script installation pattern in the CLI reference reinforces an unsafe execution path and increases the chance an agent or user will run arbitrary remote code. Duplicating it in documentation broadens exposure and normalizes risky behavior.
## CLI Reference
The `aitun` command (installed via `pip install aitun`, or alternatively `curl -fsSL https://aitun.cc/install.sh | bash` / `irm https://aitun.cc/install.ps1 | iex` on Windows) accepts these flags:
| Flag | Description |
|---|---|
The installation guidance includes shell and PowerShell one-liners that execute remote content directly, but provides no warning about the trust and integrity risks of doing so. This is dangerous because any compromise of the hosting site, transport, or script distribution path can lead to arbitrary code execution on the agent or user machine.
Using npx vite without pinning a version allows whatever version is currently resolved from the registry to be executed at runtime. That creates a supply-chain risk and reduces reproducibility, especially in an agent context where commands may run unattended and fetch fresh code from the network.
The skill instructs the user to expose a development server to the public internet but does not clearly warn that this can reveal unpublished code, debug endpoints, secrets, or unsafe dev-mode features. In context, this is especially risky because several example servers bind broadly and development servers often lack hardening appropriate for internet exposure.