Gate
Sentinel
Pre-execution checkpoint for AI agents. Lifecycle-aware verify before irreversible actions.
Product hub
Four APIs. One validation boundary.
ThoughtProof provides independent decision validation — checking whether a proposed action is justified by mandate and evidence before consequential execution. The products below are separate lines with their own contracts. They are not a single merged API.
Typical agent exit
Typical agent exit / pre-execution gate:
start with Sentinel
(POST /sentinel/verify).
Use PLV, RV, or DQL when your workflow needs plan-level, adversarial-reasoning, or five-axis checks.
Plan
PLV — Plan-Level Verification
Checks whether an agent’s plan or multi-step output still matches the stated goal and evidence — step by step, before you trust the result.
Panel
RV — Reasoning Verification
Adversarial multi-model reasoning verification with ALLOW / BLOCK / UNCERTAIN and signed attestation.
Axes
DQL — Decision Quality Layer
Scores a proposed action on five axes — intent, scope, risk, consistency, and reversibility — so weak or overreaching decisions surface before they execute.
Four separate APIs for different decision types — pick the surface that matches your workflow. Developer tooling and open-source engines are covered in the docs, not listed here as extra products.