If you discover a security vulnerability in Rampart, please report it responsibly.

Email: rampartsec@pm.me

What to include:

• Description of the vulnerability
• Steps to reproduce
• Affected version(s)
• Impact assessment (what can an attacker do?)
• Suggested fix (if you have one)

What to expect:

• Acknowledgment within 48 hours
• Initial assessment within 1 week
• Fix timeline communicated within 2 weeks
• Coordinated disclosure — we'll work with you on timing

Please do NOT:

• Open public GitHub issues for security vulnerabilities
• Disclose publicly before we've had time to respond
• Test against production systems you don't own

Rampart's threat model is documented in docs/THREAT-MODEL.md. Key design principles:

• Explicit degraded behavior — failure behavior is integration-specific and documented; managed OpenClaw protection fails closed, while compatibility paths such as preload may fail open
• Deny-wins evaluation — any deny from any policy overrides all allows
• Local-first by default — Rampart has no telemetry or mandatory cloud dependency; explicitly configured notification, decision-webhook, semantic verification, and SIEM integrations send their documented data to the operator-selected endpoint
• Hash-chained audit — tamper-evident append-only trail with SIEM export

The machine-readable integration guarantees, evidence paths, known limitations, and shared adversarial corpus live in assurance/.

The following are in scope for security reports:

• Policy engine bypasses (commands that should be denied but aren't)
• Audit trail integrity issues (undetected tampering)
• Authentication bypass on the HTTP API
• Privilege escalation through Rampart
• Information disclosure through error messages or logs

The following are out of scope:

• Denial of service against rampart serve (rate limiting is a known gap)
• Attacks requiring pre-existing root/admin access
• Social engineering
• Issues in dependencies (report upstream, but let us know)

We don't currently have a formal bug bounty program. Significant findings will be credited in the changelog and release notes (with your permission).