edgeone@1.6.38
Command-line interface for TencentCloud Pages Functions
AI Security Review
scanned 1mo ago · by lpm-firewall-aiThis is not an install-time malware chain, but it is a user-confirmed remote extension installer targeting several AI-agent control surfaces. The unauthenticated remote payload makes it a concrete extension lifecycle risk.
Decision evidence
public snapshot- The package exposes a CLI entry point that launches bundled code.
- The CLI detects home-directory AI-agent folders, prompts during Makers development, then downloads an archive from api.skillhub.cn and installs its skill files into agent skill directories.
- The downloaded archive is remote content without a pinned hash or signature, so later AI-agent use can be influenced by content not present in this package.
- package.json has no preinstall, install, or postinstall hook.
- The agent-skill installation is gated by an interactive confirmation during a user-invoked development flow.
- The source applies archive path and symlink checks and limits the download size.
Source & flagged code
14 flagged · loading sourcePackage source references child process execution.
edgeone-bin/edgeone.jsView on unpkg · L7A single source file combines environment access, network access, and code or shell execution; review context before blocking.
edgeone-bin/edgeone.jsView on unpkg · L2Findings
Affected versions and remediation
This report applies to edgeone@1.6.38.
See all versions of edgeone for other recorded verdicts.
- Review the evidence and your use of edgeone@1.6.38 before allowing it. Restrict the permissions described in this report.
- Choose an independently verified alternative or release. This report does not establish that other versions are safe.
Evidence last updated: .
Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.