@livx.cc/agentx@0.99.124
Edge-native AI agent runtime — drives a virtual filesystem via any LLM (ai.libx.js). Same bytes run in node, browser, or edge.
AI Security Review
scanned 1d ago · by lpm-firewall-aiSource inspection found no malicious behavior: package installation has no consumer lifecycle hook, and the flagged credential flow is a Bodify secrets lookup using a named Bodify credential. The scanner’s redirection and persistence claims were not substantiated by the inspected source.
Decision evidence
public snapshot- package.json defines no consumer install hook; the prepare script configures a Git hook, and the CLI main function runs when the executable is invoked.
- The Bodify lookup uses a specifically named Bodify key or vault token and sends it to the Bodify environment endpoint; it imports returned values as environment variables.
- Anthropic requests target api.anthropic.com, while the CLI passes configured API keys and optional caller configuration to the AI client.
- Inspected persistence helpers write project-local permissions or settings, and the shell and MCP files expose runtime tools and transports.
- The inspected dependency manifests and entry files show ordinary library entrypoints; no attack behavior was established in those required source locations.
- The scanner’s credential-redirection and persistence labels are not supported by the inspected source: no fixed unofficial AI gateway or automatic credential-redirection write was found.
Source & flagged code
13 flagged · loading sourceThis package version adds a dangerous source file absent from the previous stored version; route for source-aware review.
dist/tools.shell.jsView on unpkgPackage source references child process execution.
dist/tools.shell.jsView on unpkg · L891Findings
Affected versions and remediation
This report applies to @livx.cc/agentx@0.99.124.
See all versions of @livx.cc/agentx for other recorded verdicts.
- Review the evidence and your use of @livx.cc/agentx@0.99.124 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.