agentiqa@1.1.77
AI-powered testing for web and mobile apps
AI Security Review
scanned 26d ago · by lpm-firewall-aiThe package has an install-time Node-version gate and a CLI that can route authenticated runs to its hosted engine or execute locally. Source inspected here does not substantiate a concrete attack, but the remote agent capability remains a warning-level risk.
Decision evidence
public snapshot- The package runs a preinstall hook that checks the Node version.
- The CLI reads provider API keys from the environment and supports a configurable OpenAI base URL.
- The embedded engine exposes HTTP and WebSocket interfaces; non-health requests require the configured engine token when one is set.
- Authenticated use selects a hosted Agentiqa engine by default, enabling remote agent execution.
- The preinstall script only checks the Node version and exits on unsupported versions.
- OpenAI provider calls use the configured base URL or the standard OpenAI endpoint.
- The CLI exposes an embedded execution option, and the hosted engine URL is within the package vendor's domain.
Source & flagged code
8 flaggedPackage defines install-time lifecycle scripts.
package.jsonView on unpkgInstall-time lifecycle script is not statically allowlisted and needs review.
package.jsonView on unpkgFindings
Affected versions and remediation
This report applies to agentiqa@1.1.77.
See all versions of agentiqa for other recorded verdicts.
- Review the evidence and your use of agentiqa@1.1.77 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.