claude-codex-wechat@0.1.60
用微信,遥控你自己电脑上的 `Claude Code` 和 `Codex CLI`。
AI Security Review
scanned 9d ago · by lpm-firewall-aiThe package has no install hook, but a user-started daemon binds to all interfaces and exposes management routes that disclose local state and directory names without handler-level authentication. This warrants a warning for network-exposed capability; the inspected source does not prove a block-eligible attack.
Decision evidence
public snapshot- The manifest exposes the server CLI and defines publish-time scripts; running it without a subcommand starts the daemon.
- The daemon defaults to listening on all network interfaces at port 8787.
- The directory-listing route accepts a caller-supplied path and enumerates its directories without an authentication check in the handler.
- The state route returns active-user and runtime configuration data, while session routes accept caller-supplied user and session identifiers.
- On a Claude session-attach request, the runtime checks an idle background-session record and may delete its recorded path; this is user-requested session handling, not install-time activity.
- The manifest has no install hook; the daemon is started by invoking the CLI.
- The reported session cleanup is tied to a session-attach request, and the Douyin downloader is a separately invoked CLI script.
Source & flagged code
7 flagged · loading sourceSource fetches a remote non-code asset, decodes its contents, and dynamically executes the decoded payload.
dist/server/cli.jsView on unpkg · L4A package entrypoint or install-time lifecycle script reaches a source file with blocking dangerous behavior.
dist/server/cli.jsView on unpkgFindings
Affected versions and remediation
This report applies to claude-codex-wechat@0.1.60.
See all versions of claude-codex-wechat for other recorded verdicts.
- Review the evidence and your use of claude-codex-wechat@0.1.60 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.