What We Build
Products
Remnant Fieldworks builds verification infrastructure for consequential autonomous systems. ExecutionProof is the company's primary product: a pre-execution verification and control layer that decides whether a consequential action should be allowed before it executes. Two further research directions extend the same principle to whether current state can be trusted (StateProof) and whether a system that lost trust is safe to resume (ReEntryProof).
ExecutionProof™ - Pre-Execution Authorization
Live, pilot-readyExecutionProof™ is the primary Remnant Fieldworks product. It is a pre-execution verification and control layer for consequential AI actions. Before any consequential action executes, ExecutionProof checks authority, evidence, policy constraints, and current system state, then returns ALLOW, HOLD, or DENY. Every decision is recorded in a signed ProofRecord™.
Visit executionproof.ioStateProof - State Trustworthiness Verification
Research, simulation-stageStateProof is a research direction, not a current commercial product. It addresses whether current system state is sufficiently trustworthy for consequential action. One active application domain is low-altitude air domain awareness. Current status: simulation-stage research, TRL 3.
ReEntryProof - Trusted System Recovery
Under investigationReEntryProof is a future research direction, not a current commercial product. It addresses what must be proven before a system that lost trust is allowed to resume, sitting after ExecutionProof returns HOLD or after StateProof declares state untrustworthy. Current status: conceptual, under investigation, not yet built.
VaultProof Agent Guard is available as a free, open-source pre-execution verification tool. View the VaultProof Agent Guard repository.
RF-100 is the draft verification governance standard (v0.9, public review). ExecutionProof™ is RF’s primary commercial implementation of pre-execution verification principles. View the RF-100 repository.