Methodology & ethics

Purpose: pathfinding research

Frontier Labs practices anticipatory threat research: model a plausible future threat class end to end, in the open, so that detection engineering can evolve before the threat arrives, not after. The question is never "how would you build this" but "what would you still see if someone did."

The flagship track, Novich0k, applies that method to a specific forecast: command-and-control infrastructure adopting zero-trust patterns, with cryptographically blind, threshold-split, disposable relays.

Scope discipline

• Simulation only. Every interactive element on this site runs on mock data with no functional backend. Nothing here can issue a command to anything.
• No deployable artifacts. No code, build instructions, configurations, or operational parameters are published or linked. The single exception to "nothing to download" is text-based hunt notes exported from the simulation page.
• Detection counterpart rule. No capability is described without its defensive counterpart on the same page.
• Confidence discipline. Where detection feasibility is uncertain, the site says so. No efficacy claims are invented in either direction.
• Abstraction limits. Sensitive mechanics are described at detection-relevant depth only: what traffic looks like and how to tell it apart, never byte-level recipes.

Authorization & ethics

This is independent research. No third-party systems, networks, or data are touched at any point. The cryptography discussed (Umbral proxy re-encryption on secp256k1) is published, peer-adjacent work available as open libraries; this site builds arguments on top of it, not exploits against anyone. Third-party frameworks named on this site (Cobalt Strike, Havoc, Sliver, Mythic) are discussed factually, citing only their public, documented behavior.

Author

Rhys Evans is a threat researcher focused on anticipating how offensive infrastructure evolves, and on making that analysis useful to the defenders who have to see it coming. Frontier Labs is the public record of that work.

Contact and writing: Medium · GitHub

Responsible disclosure

If you believe something published here creates risk that its framing fails to mitigate, that is exactly the conversation this page exists for. Reach out via Medium with the specific claim or artifact and the concern; corrections and retractions will be made publicly and promptly.