@jait/gateway@0.1.946
⚠ Under reviewJait AI gateway — local-first AI coding agent with terminal, filesystem, and browser control
AI Security Review
scanned 5d ago · by lpm-firewall-aiInspected source did not confirm malicious behavior: the credential finding maps to a configurable web search client, and the agent configuration is session-scoped. Manual review is warranted because the required clean citation coverage exceeds the 128-reference limit; the automatic postinstall runtime provisioning is not, by itself, evidence of an attack.
Decision evidence
public snapshot- The postinstall hook runs the runtime installer, which provisions a pinned Graphify package in a Python virtual environment when needed.
- The web search tool uses configured provider keys and an optionally configured OpenAI-compatible base URL when the tool is invoked.
- The Pi integration creates a per-session MCP config instead of changing global or project configuration.
- Inspected source did not establish credential theft or another concrete attack; the flagged credential flow is part of the configurable web search client.
- The Pi integration keeps its MCP configuration scoped to a session.
Source & flagged code
30 flagged · loading sourcePackage defines install-time lifecycle scripts.
package.jsonView on unpkgPackage source references child process execution.
bin/safe-update.mjsView on unpkg · L1Findings
Affected versions and remediation
This report applies to @jait/gateway@0.1.946.
See all versions of @jait/gateway for other recorded verdicts.
- Review the evidence and your use of @jait/gateway@0.1.946 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.