devez-vibe@1.9.35
Stable terminal UI for Codex and Claude Agent SDK
AI Security Review
scanned 24d ago · by lpm-firewall-aiThe automatic cross-agent skill installation and user-level pip mutation create a guarded agent-extension lifecycle risk. The inspected source does not establish a concrete malicious attack, so blocking is not supported.
Decision evidence
public snapshot- The manifest automatically runs the skill installer and dependency preparer after npm installation.
- The installer targets Codex and Claude skill directories under the user's configuration homes and copies, prunes, and marks two bundled skill directories.
- The same automatic lifecycle chain invokes a user-level pip installation when its Python dependency check fails.
- Claude permission updates are exposed through a bridge function and default to project settings; this behavior is separate from the install hook.
- No inspected source showed secret collection, credential forwarding, remote payload retrieval, or hidden network exfiltration.
- The installed agent content is limited to package-named skill directories rather than arbitrary agent configuration files.
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@1.9.35.
See all versions of devez-vibe for other recorded verdicts.
- Review the evidence and your use of devez-vibe@1.9.35 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.