Launching today
ExecutionProof™
Stop AI Agents From Moving Money Without Proof
5 followers
Stop AI Agents From Moving Money Without Proof
5 followers
Remnant Fieldworks builds ExecutionProof — the verification layer that stops AI agents from moving money without proof. We gate vendor and treasury payments before they execute, returning ALLOW, HOLD, or DENY with a signed ProofRecord.
Every autonomy conversation eventually arrives at the same problem: who authorized this?
When an AI agent initiates a payment, approves a vendor, or modifies a treasury record, the traditional answer is: check the log afterward. That is not authorization. That is forensics.
ExecutionProof is the verification layer built before the action. Before a consequential step executes, ExecutionProof evaluates whether the agent has verified authority, authenticated evidence, a scoped policy snapshot, and declared system state — and returns ALLOW, HOLD, or DENY. The decision and its basis are preserved in a ProofRecord before execution proceeds.
We built RF-100 — an open standard for pre-execution control — because no existing framework specified this gate. RF-100 maps to OWASP Agentic AI Top 10 and NIST AI RMF, and is available for public review at github.com/derekhone/rf-100. ExecutionProof is the pilot-ready implementation.
This is a new category: Verification Before Execution™. Not observability. Not audit. The gate itself.
56 patent applications pending. We are looking for 3–5 design-partner pilots — teams building autonomous payment or treasury workflows who want to implement pre-execution control with shared input into the standard.
If this is on your radar: https://calendly.com/derek-owner...