01 AI agent sessions
AgentProof
Local, hash-only receipts of collector-observed agent runs and checked repository commitments.
Codex + OpenClaw available · Claude Code under review
Explore AgentProofCurrent work
Reviewable evidence for AI and real operations.
Two local AI plugins, one independent evidence core and a live property-management vertical.
HREVN preserves narrow, checkable facts. Integrity is not the same as truth, completeness, authority or approval.
See the four layers 02 Signed AI evaluations
EvalDossier
Offline verification of supported signed evaluation dossiers, without collapsing formal checks, model judgments or inconclusive outcomes.
Codex + OpenClaw available · Claude Code under review
Explore EvalDossier 03 The technical heart
HREVN EB1 Core
A standalone, product-neutral core for deterministic evidence bundles and strict, multi-axis verification.
EB1 v2 · No single global “valid” result
Understand Evidence Bundles 04 A vertical in operation
SuperProperty
AI-ready property operations with governed read-only queries, portable data exports and native EB1 v2 evidence bundles.
English product landing · Live synthetic demonstration
Explore SuperPropertyOne evidence discipline
Observe. Verify. Package. Apply.
Different products, one rule: state exactly what was checked and keep everything else visible.