devez-vibe@1.10.6
Stable terminal UI for Codex and Claude Agent SDK
AI Security Review
scanned 6d ago · by lpm-firewall-aiThe postinstall hook performs first-party setup of skills in global agent directories, which leaves a concrete agent-extension lifecycle risk and merits a warning. The inspected skill text and bridge behavior do not establish a concrete attack or justify blocking publication.
Decision evidence
public snapshot- The postinstall hook automatically runs the skill installer and dependency preparer.
- The installer copies and prunes bundled skills in the user’s Codex and Claude skill directories.
- The installed skills direct agents to run bundled workflows, including an adaptive web-fetch tool.
- The Luna skill says to use it only when the user directly invokes it; the bridge exposes permission updates through a runtime request handler.
- The inspected bridge permission-setting path is a runtime operation, not part of the postinstall installer.
- The bundled Luna skill specifies direct user invocation as its activation condition.
Source & flagged code
8 flagged · loading sourcePackage defines install-time lifecycle scripts.
package.jsonView on unpkgInstall-time lifecycle script is not statically allowlisted and needs review.
package.jsonView on unpkgSource creates an unconsented AI-agent control surface through install-time mutation or a default unauthenticated remote skill channel.
Findings
Affected versions and remediation
This report applies to devez-vibe@1.10.6.
See all versions of devez-vibe for other recorded verdicts.
- Review the evidence and your use of devez-vibe@1.10.6 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.