@hmharness/cli@0.23.33
hmharness command line: one-shot tasks, an interactive REPL, a fullscreen TUI, the web frontend, and direct tool invocation.
AI Security Review
scanned 17d ago · by lpm-firewall-aiSource inspection found a conditional command-execution risk in the automatic update fallback: model-proposed commands are incompletely validated before Windows shell execution. The package has no install lifecycle hook, and the inspected behavior does not establish a concrete malicious payload, so a warning is appropriate.
Decision evidence
public snapshot- The manifest has no install lifecycle hook and lists matching-version first-party dependencies.
- The REPL starts automatic updates by default; setting its update option to false changes this to notification only.
- When an automatic update fails, the updater asks the configured model for an installation command using machine details and the error.
- The command check rejects selected dangerous words and requires a reference to this package, but does not reject shell operators.
- On Windows, the accepted model response is passed as a command string to the system command interpreter.
- The risky path requires a failed automatic update and a model response that passes the package-name check.
- The update feature is scoped to installing this CLI package; the inspected source does not establish a malicious payload or credential transfer.
Source & flagged code
5 flaggedPackage source references child process execution.
board/install-kaihongos.cjsView on unpkg · L35A single source file combines environment access, network access, and code or shell execution; review context before blocking.
board/install-kaihongos.cjsView on unpkg · L339Findings
Affected versions and remediation
This report applies to @hmharness/cli@0.23.33.
See all versions of @hmharness/cli for other recorded verdicts.
- Review the evidence and your use of @hmharness/cli@0.23.33 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.