AuthContract turns institutional rules into developer-usable, testable controls. It connects authoritative rule sources to canonical contracts, CI/CD merge gates, runtime action checks, and reconstructable evidence—so teams can prove not just what an agent did, but which rule authorized it, what version applied, and why execution was allowed.

auditabilityauthorizationauthorization-policyci-cddecision-evidence

LEGION

Repo · 2 stars · Python

▲ 0

LEGION-100: Verified multi-agent execution infrastructure | 100 agents, 13 executable worlds, CAGE + ZTL + VEIP, independent Judge and Verified Collective Completion VCC@1

agent-orchestrationautonomous-agentsenterprise-aiformal-methodsformal-verification

LEGION-Federation

Repo · 1 star · Python

▲ 0

Open-source infrastructure connecting independent AI agent collectives. Enables secure discovery, controlled information sharing, multi-party negotiation, verifiable contracts, and coordinated execution through a shared LEGION Core.

agent-federationagent-to-agentagent2agentcommercedistributed-systemsformal-verification

RAED

Repo · 1 star · Go

▲ 0

100K-Agent / 1M-Connection Persistent Machine Society

Veraxis-Memory-Admissibility-Management-MAM-

Repo · 2 stars · Go

▲ 0

Veraxis MAM is a zero-allocation, register-speed memory admissibility firewall for multi-agent runtimes. Evaluates context validation in <34ns with 0 B/op. Intercepts indirect prompt injections, detects out-of-band subspace drift, and seals cryptographic lineage. Governing memory before context, and context before inference.

About

Arkadiy Miteiko is the Founder of Veraxis, a company that builds open infrastructure for Open Institutional Computation, which keeps human-written meaning, warrant and authority attached to what machines do. Arkadiy publishes open-source work on GitHub, including AuthContract, which turns institutional rules into testable controls, and LEGION, a multi-agent execution infrastructure. He is focused on verifiable execution and the evidence that lets teams see which rule authorized an agent's action.