@livx.cc/agentx@0.99.92
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 4d ago · by lpm-firewall-aiThe reviewed behaviors match the package’s agent runtime and vault integration. The source does not establish credential redirection, unauthorized agent configuration mutation, or another concrete attack.
Decision evidence
public snapshot- package.json exposes a CLI; its prepare script sets a Git hooks path, and prepublishOnly builds the package.
- The flagged CLI location constructs a .claude/worktrees path for isolated Git worktrees.
- The runtime loads project instruction files and appends them to the prompt when instruction loading is enabled; this is agent runtime behavior.
- The declaration describes typed child-agent configuration, including an optional system prompt.
- Shell and MCP clients spawn configured processes as runtime tools.
- The CLI uses a Bodify vault token to request environment values from the Bodify API, then makes returned values available to the runtime.
Source & flagged code
12 flagged · loading sourcePackage source references child process execution.
dist/tools.shell.jsView on unpkg · L891Findings
Affected versions and remediation
This report applies to @livx.cc/agentx@0.99.92.
See all versions of @livx.cc/agentx for other recorded verdicts.
- Review the evidence and your use of @livx.cc/agentx@0.99.92 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.