cartethyia@2.3.2
A powerful, self-hosted reverse proxy for OpenAI- and Anthropic-compatible AI clients, powered by native Bun and Elysia. It serves both protocols from a single server, routes every model across any configured provider, and translates requests and response
AI Security Review
scanned 5d ago · by lpm-firewall-aiSource inspection did not substantiate the scanner’s credential-exfiltration or runtime-install claims: the CLI’s external request checks version metadata, and its package update is user-triggered. No install lifecycle hook or concrete attack behavior was identified.
Decision evidence
public snapshot- The manifest exposes command-line entrypoints and shows development, packaging, and setup scripts; it has no preinstall, install, or postinstall hook.
- The inspected CLI uses the npm registry for version checks and runs a global update only after a menu or tray update action. Its other inspected network calls target the local gateway.
- The flagged catalog file describes a generated offline model catalog. The cited runtime and support paths identify gateway, routing, authentication, provider, and observability components; the inspected excerpts establish no attack behavior.
- Scanner labels alleged credential exfiltration and runtime installation, but inspected CLI network requests check package version or contact the local gateway; the global package update is tied to a user-selected update action.
- No concrete credential transfer, install-time mutation, or malicious behavior was established in the inspected sources.
Source & flagged code
12 flagged · loading sourcePackage contains a high-severity secret pattern.
src/providers/discovery/catalog/catalog-blob.tsView on unpkg · L15Google API key in src/providers/discovery/catalog/catalog-blob.ts
src/providers/discovery/catalog/catalog-blob.tsView on unpkg · L15Findings
Affected versions and remediation
This report applies to cartethyia@2.3.2.
See all versions of cartethyia for other recorded verdicts.
- Review the evidence and your use of cartethyia@2.3.2 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.