Product thesis
The next security layer is not merely login; it is proving who or what acted, under which authority, through which session and transport, under which policy, with evidence that can be verified later.
2026 – Present
TrustForge
Open-source trust fabric for AI-native software, secure devices, authenticated sessions, explicit authority, policy, and verifiable action.
- Open source
- AI security
- TypeScript
- Rust
- Protocol design
Results at a glance
Results that mattered
A snapshot of the results and changes across operations, adoption, and team experience.
- Published crates
- 31
- 4,298 combined crates.io downloads as of July 18, 2026.
- npm reach
- 6.5K+
- Trailing-year downloads across ten verified packages as of July 18, 2026.
- Source scale
- 109K
- Tracked source lines under the portfolio measurement method.
Approach and execution
What I did
Decisions I directed
The decision record covers live and packet modes, transport migration, relay identity, constrained networks, break-glass authority, actor and model identity, agent contracts, proof storage, policy, approval ceremonies, delegation, revocation, and conformance.
AI-native implementation
AI coding agents generated the implementation. Kody owned product direction, requirements, service selection, infrastructure, investigation, acceptance, release decisions, and production outcomes.
Current status
TrustForge is explicitly experimental, not production-ready. Published packages and working references demonstrate breadth; hardware coverage and hardening remain uneven and are labeled honestly in the project itself.
Interested in similar results?
I’m available for full-time roles or contract engagements—let’s discuss how my hands-on approach can help your team.