AI Security Review
scanned 9d ago · by lpm-firewall-aiNo confirmed malicious attack surface was found. The package is a security scanner/remediation CLI with user-invoked network, filesystem patching, git hook, and MCP setup features.
Decision evidence
public snapshot- src/cli/index.ts and index.js include explicit install-mcp commands that write AI-client MCP config files.
- src/telemetry/cloudReporter.js can POST scan findings and repo metadata when a license key and findings exist.
- src/router/cloud.js, src/remediation/patcher.ts, and src/release-gate/patcher.ts send source/diffs to PreFlight remediation endpoints for requested fixes.
- package.json has no install/preinstall/postinstall lifecycle hooks.
- cli.js only loads src/cli/index.ts main; no import-time payload beyond CLI dispatch.
- MCP config writes are behind explicit install/init commands, not install-time or hidden execution.
- Network endpoints are product-aligned: licensing, login, telemetry, and remediation proxy.
- No credential harvesting, cloud metadata request, persistence, destructive payload, or dependency confusion behavior confirmed in source.
- child_process use is package-aligned: git hook setup, npm audit, browser open, clipboard, daemon/legacy CLI, and hardware probe.
Source & flagged code
8 flagged · loading sourceThis package version adds a dangerous source file absent from the previous stored version.
index.jsView on unpkgA single source file combines environment access, network access, and code or shell execution; review context before blocking.
index.jsView on unpkg · L3Source reaches cloud instance metadata or link-local credential endpoints.
index.jsView on unpkg · L3Package source references weak cryptographic algorithms.
src/daemon/protocol.tsView on unpkg · L57Source gates dangerous network, credential, or execution behavior behind CI, host, platform, time, or geo fingerprint checks.
src/cli/login.jsView on unpkg · L1