@1interface/voice-core@0.1.0
AI Security Review
scanned 4mo ago · by lpm-firewall-aiThe install hook is unusual but scoped to package-local dependency cleanup, and the runtime capabilities match a voice-core package without hardcoded exfiltration or persistence. No concrete malicious behavior or unconsented AI-agent control mutation was found by source inspection.
Decision evidence
public snapshot- package.json has postinstall: rm -rf node_modules/react node_modules/react-dom
- dist runtime opens WebSocket from @1interface/shared-core API_ENDPOINTS.CONVERSATION_STREAM and sends audio/text after voice connect
- dist runtime requests microphone permission via navigator.mediaDevices.getUserMedia
- No AI-agent control-surface writes or home/project persistence found
- No child_process, eval, fs writes, credential harvesting, or hardcoded exfiltration host found in dist files
- Postinstall only removes package-local nested React copies, consistent with peer dependency cleanup
- Network/microphone behavior is user-invoked voice-chat functionality and package-aligned
- Published files are limited to dist plus package manifest
Source & flagged code
2 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 unpkgFindings
Affected versions and remediation
This report applies to @1interface/voice-core@0.1.0.
See all versions of @1interface/voice-core for other recorded verdicts.
- Review the evidence and your use of @1interface/voice-core@0.1.0 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.