Skip to content

Security: airblackbox/gateway

Security

SECURITY.md

Security Policy

Reporting Vulnerabilities

If you discover a security vulnerability in AIR Blackbox Gateway, please report it responsibly.

Email: jason.j.shotwell@gmail.com Subject line: [SECURITY] AIR Blackbox Gateway — <brief description>

We will acknowledge your report within 48 hours and aim to provide a fix or mitigation within 7 days for critical issues.

Please do not open a public GitHub issue for security vulnerabilities.

What Data AIR Blackbox Gateway Stores

The gateway records AI system interactions. Here is exactly what is stored and where:

Data Where It Goes Who Controls It
Raw prompts & completions Your MinIO/S3 vault (never leaves your infrastructure) You
Vault references (URIs, not content) OTel traces → your collector → your Jaeger/Grafana You
Model name, token counts, timing OTel span attributes You
Run ID, trace ID, timestamps .air.json record files on your filesystem You
SHA-256 checksums of request/response .air.json record files You

What AIR Blackbox Gateway Does NOT Do

  • No phone-home. The gateway makes zero network calls except to your configured upstream LLM provider and your own infrastructure (MinIO, OTel Collector).
  • No telemetry. We do not collect usage data, crash reports, or analytics.
  • No cloud dependency. Everything runs on your infrastructure. There is no SaaS component.
  • No content in traces. OTel spans contain vault references, not raw prompts or completions. Your observability stack never sees sensitive content.
  • No credential storage. API keys are passed through to the upstream provider and are not written to AIR records or vault storage.

Threat Model

AIR Blackbox Gateway sits in the request path between your AI agent and the LLM provider. The primary security considerations are:

  1. Vault access — MinIO/S3 credentials control access to stored prompts and completions. Protect these credentials with the same rigor as your LLM API keys.
  2. AIR record files — These contain vault references and checksums, not raw content. However, metadata (model names, timestamps, token counts) may still be sensitive in some contexts. Apply appropriate filesystem permissions.
  3. Network position — The gateway terminates your agent's API call and forwards it. It sees the full request and response in transit. Deploy it in the same trust boundary as your agent.

Supported Versions

Version Supported
Latest on main Yes
Older commits Best effort

Responsible Disclosure

We follow coordinated disclosure. If you report a vulnerability, we will:

  1. Acknowledge receipt within 48 hours
  2. Provide an estimated timeline for a fix
  3. Credit you in the release notes (unless you prefer anonymity)
  4. Not pursue legal action against good-faith security researchers

There aren’t any published security advisories