Skip to content

Security: BalianWang/busytok

Security

SECURITY.md

Security Policy

Supported versions

Busytok is pre-1.0. Only the latest minor release receives security fixes. The 0.x line explicitly does not promise backports — see README for the 0.x stability contract.

Reporting a vulnerability

Preferred channel: GitHub Private Vulnerability Reporting.

Open https://github.com/BalianWang/busytok/security/advisories/new and submit a private advisory. This routes directly to the maintainer without public disclosure.

If GitHub PVR is unavailable, email the maintainer directly (see the GitHub profile). Do not open a public GitHub issue for security reports.

Please include:

  • Busytok version (from Settings → About, or busytok --version)
  • macOS version
  • The agent type whose logs were being audited (Claude Code / Codex / Gemini CLI)
  • The log path that triggered the issue (if applicable)
  • Reproduction steps
  • Impact assessment

Response timelines

  • Acknowledgement: within 72 hours
  • Initial assessment: within 7 days
  • Fix or mitigation: target 30 days for high-severity issues, 90 days for low-severity

These are targets, not SLAs — this is a one-maintainer project.

Scope

Busytok is a local-first agent token audit dashboard. It reads local AI agent log files, persists metadata-only token usage to local SQLite, and exposes a local GUI/CLI. It does not proxy network traffic, store credentials, modify client configurations, or handle provider OAuth/session tokens.

Vulnerabilities in any of these properties are in scope:

  • Sandbox escape via the GUI (webview → filesystem → user session)
  • Privilege escalation via the bundled busytok-service (LaunchAgent / SMAppService)
  • SQLite injection via crafted log content
  • Crash loops triggered by malformed agent logs that prevent the app from starting
  • Path traversal in log path discovery

Out of scope:

  • Vulnerabilities in third-party agents themselves (report to the agent's maintainer)
  • Vulnerabilities in dependencies — these are tracked by cargo-audit and dependabot; report directly to the upstream project

Triage scope note

Busytok reads local AI agent logs. A reporter may include log excerpts that contain prompt content. The maintainer will treat all such material as confidential and delete it after the report is resolved.

There aren't any published security advisories