@grovetech/defender@0.32.1
Active runtime protection for vibe-coded apps — drops in as Express middleware (web + AI layer), blocks prompt injection, PII leaks, and sensitive paths in real time. By Grovetech AI.
AI Security Review
scanned 3d ago · by lpm-firewall-aiThe SDK enables runtime telemetry by default when an API key is present and sends identifying host and request information to the vendor endpoint. This is suspicious telemetry, not evidence of credential exfiltration; the CLI credential flow is tied to an explicit replay command and the package’s own service.
Decision evidence
public snapshot- With an API key configured, the SDK enables telemetry unless `telemetry: false` is set.
- Telemetry POSTs the host name, package version, request count, and queued events.
- WAF violation events can include a raw client IP address, request path, and excerpt.
- The default telemetry endpoint is `https://grovetechai.com/api/defender/ingest`.
- The package’s middleware purpose includes reporting violations to telemetry.
- The telemetry is sent to the package vendor’s own service and carries the configured Defender API key as authorization.
- Without an API key, the SDK says it runs in audit-only mode with no telemetry uploads.
- The flagged CLI replay flow requires an explicit `replay` command and sends the configured Defender key to the package’s own API.
Source & flagged code
8 flaggedPackage text addresses the security reviewer or scanner and tries to influence the review outcome.
README.mdView on unpkgSource appears to send environment or credential material to an external endpoint.
dist/esm/cli.jsView on unpkg · L4Findings
Affected versions and remediation
This report applies to @grovetech/defender@0.32.1.
See all versions of @grovetech/defender for other recorded verdicts.
- Review the evidence and your use of @grovetech/defender@0.32.1 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.