registry / @golproductions/check / 4.3.2

@golproductions/check@4.3.2

Anti-hallucination layer for Claude Code: before a command or file write runs, your own machine checks it, and what it proves missing is blocked.

AI Security Review

scanned 1d ago · by lpm-firewall-ai

No concrete attack was identified. The inspected network and subprocess behavior is consistent with the package's command-checking purpose, though obfuscation limits verification.

Static reason

One or more suspicious static signals were detected.

Trigger

Running the package's command-line entrypoint for a check.

Impact

No malicious effect was established from the inspected source.

Mechanism

It sends a machine fingerprint and client ID in service requests, then runs bounded shell probes and parses their results.

AI rationale

The obfuscated entrypoint uses network and subprocess capabilities, but the inspected request and probe flow is consistent with the stated Claude Code checking function. The package has no install lifecycle hook, and the inspected source did not establish credential theft, unconsented agent-control mutation, or another concrete attack.

Decision evidence

public snapshot

AI called this Clean at 77.0% confidence as Benign with medium false-positive risk.

Evidence for AI clean decision

The package describes itself as a Claude Code command and file-write checking tool.
Its prepublishOnly hook runs build and version/upgrade checks; no install lifecycle hook is declared.
The entrypoint sends a machine fingerprint to a preflight route and makes authenticated requests using a GOL client ID.

@golproductions/check@4.3.2

AI Security Review

Decision evidence

Source & flagged code

Source & flagged code

Findings

Affected versions and remediation

Source & flagged code