# Security Architects, Inc

> Security products and consulting since 1999. Founded in San Diego,
> California. We build hardened, open-source security tools — with a current
> focus on AI agent security — and provide professional services from
> penetration testing to AI red-teaming. Philosophy: enforce good behaviour
> instead of chasing bad behaviour; default-deny, explicit allow.

## Products

- [Cone of Silence](https://securityarchitects.com/products/cone-of-silence/):
  Agent Pods — hardened, always-on, sandboxed pods for Claude and other AI
  agents, reachable via Telegram. Managed service: OS-level sandbox isolation
  by default, per-project scoped credentials, prompt-injection scanning,
  paired-device access. From $195/month plus $1,500 onboarding.
- [Claudette](https://securityarchitects.com/products/claudette/): open-source
  (ISC) hardened sandbox for running AI coding agents such as Claude Code —
  bubblewrap isolation, per-project SSH keys, prompt-injection defense.
  Source: https://gitlab.com/jowolf/claudette
- [Authinator3](https://securityarchitects.com/products/authinator/):
  open-source (ISC) web login gateway for firewalls — authenticate over
  HTTPS and your IP is granted passage through OpenBSD pf or Linux nftables.
  Better than port knocking. Source: https://gitlab.com/jowolf/authinator3
- [Truepost](https://securityarchitects.com/products/truepost/): email client
  that feels like messaging, with sender identity treated as a security
  surface. Public beta planned at truepost.com.
- [1lan](https://securityarchitects.com/products/1lan/): reach machines
  behind NAT and firewalls through your own relay using plain OpenSSH —
  self-hosted alternative to commercial tunnel services.
- [Ozone (legacy)](https://securityarchitects.com/products/ozone/): our
  1999–2007 Host Intrusion Prevention System for Windows — least-privilege
  enforcement two decades before "zero trust." Kernel source released free
  in 2004.

## Services

- [Services overview](https://securityarchitects.com/services/): AI security
  (red-teaming, LLM penetration testing, agent sandbox security, AI
  governance/NIST AI RMF/EU AI Act, ML supply-chain review, threat modeling
  vs MITRE ATLAS) and core security services (penetration testing, code
  review, hardening, incident response, training). Scoping call $500,
  credited toward any engagement; day rates from $3,500.

## Company

- [About](https://securityarchitects.com/about/): founded 1999, San Diego.
- [Contact](https://securityarchitects.com/contact/): info@securityarchitects.com
- [News](https://securityarchitects.com/post/): announcements and research
  commentary.
- [Vintage site](https://securityarchitects.com/vintage/): the original
  2000s secarch.com, preserved for historical interest.