@inboost-pro/inboost-ai-proxy@0.3.3
Local inference proxy and reliability gateway for Claude Code, Cursor, OpenHands, and Cline
AI Security Review
scanned 4d ago · by lpm-firewall-aiThe lifecycle performs automatic first-party agent setup across global and project configuration files, creating a residual agent extension risk. The inspected behavior supports a warning, but does not establish credential theft or another concrete attack.
Decision evidence
public snapshot- The package postinstall script runs the CLI with its postinstall flag.
- That install path calls the agent setup routine, which writes global and project agent files across several tools.
- The setup adds proxy and workflow instructions to those files, changing agent behavior without a setup prompt.
- The postinstall handler exits after setup, and the inspected path does not start the proxy service or send credentials.
- The added instructions configure the package's local proxy and describe editing and verification practices; the inspected lifecycle path shows no credential transfer.
Source & flagged code
12 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 @inboost-pro/inboost-ai-proxy@0.3.3.
See all versions of @inboost-pro/inboost-ai-proxy for other recorded verdicts.
- Review the evidence and your use of @inboost-pro/inboost-ai-proxy@0.3.3 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.