@yhong91/vibetime@0.1.82
vibetime CLI — install AI-agent hooks (Claude Code, Codex, OpenCode, Pi, Cursor) and report activity to vibetime.
AI Security Review
scanned 13d ago · by lpm-firewall-aiThe bundle sends a configured API token to a fixed IP endpoint over unencrypted HTTP by default, creating a concrete conditional credential exposure risk. The user-run CLI, explicit token configuration, and absence of an install hook do not establish a broader attack chain.
Decision evidence
public snapshot- The manifest exposes the bundle as a user-run CLI and declares no install lifecycle hook.
- The CLI’s default API URL is an IP address using unencrypted HTTP; an environment or stored URL can override it.
- The client attaches the saved API token as a bearer credential to requests sent to the selected API URL.
- The ingest command sends activity rollups to that URL, creating a conditional credential exposure risk.
- Agent integration files are written through the explicit install command.
- The updater invokes a global npm install only when the user runs the upgrade command.
- The token is explicitly set or supplied through CLI or environment configuration; the inspected code does not show harvesting of unrelated service credentials.
- The CLI supports a user-selected API URL, and its agent setup is user-invoked rather than an install hook.
Source & flagged code
5 flagged · loading sourceFindings
Affected versions and remediation
This report applies to @yhong91/vibetime@0.1.82.
See all versions of @yhong91/vibetime for other recorded verdicts.
- Review the evidence and your use of @yhong91/vibetime@0.1.82 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.