Codex Analytics Ledger reads sensitive local Codex metadata. Security reports must never include real prompts, transcripts, databases, authentication files, environment variables, usernames, or absolute machine paths.

Until the first stable release, security fixes are provided only for the latest published 0.x version.

Use GitHub's private vulnerability reporting flow under Security → Report a vulnerability. Repository maintainers should enable private vulnerability reporting before the first public release.

If private reporting is unavailable, contact a maintainer privately through their GitHub profile and request a secure reporting channel. Do not open a public issue containing exploit details or sensitive data.

Include only sanitized information:

• A concise description and realistic impact
• Affected version and commit
• Operating system, Node.js version, and Codex version
• Minimal reproduction steps using synthetic data
• Suggested mitigation, when known

Maintainers should acknowledge a complete report within seven days and provide status updates as investigation progresses. Disclosure timing will be coordinated with the reporter after a fix or mitigation is available.

The supported default assumes:

• The standard scripts are used, preserving the 127.0.0.1 binding.
• The machine and local user account are trusted.
• Codex state is readable only by intended local users.
• The dashboard is not placed behind a public proxy or port-forward.

Remote deployment, shared-machine isolation, and protection from a compromised local account are outside the current security boundary.

Never commit or attach:

• .codex/ directories
• auth.json, configuration secrets, or access tokens
• state_*.sqlite, *.db, journal, WAL, or shared-memory files
• Session or archived-session JSONL files
• .env files
• Screenshots containing prompts, project paths, usernames, or customer data

Use synthetic fixtures for security research and compatibility tests.