On this page
Structured tool access for AI agents, with authorization, write control, and audit kept firmly on the server.
Research status: Preprint. Original research date: 2026-02-22.
Authors: Genliang Zhu, Chu Wang, Ziyuan Wang, Zhida Li, Qiang Li.
Overview
AI agents need more than a list of callable tools. They need a server-side boundary that decides what a credential may discover, what it may read, and how proposed writes are reviewed and executed.
OpenPort defines that boundary through authorization-dependent discovery, scoped and attribute-based policy checks, draft-first writes, high-risk preflight and replay safeguards, stable machine-readable failures, rate limits, and structured audit events. A reference runtime and conformance tooling make these governance properties executable and testable across models and agent runtimes.
Core contributions
Protocol contract
Defines a model- and runtime-neutral interface for tool discovery and invocation, including endpoints, response envelopes, tool metadata, and stable agent.* reason codes.
Governed execution
Combines integration credentials, scopes, and ABAC-style constraints with draft-first writes, preflight impact binding, idempotency, optional State Witness checks, rate limits, and structured audit.
Executable verification
Pairs the specification with a reference runtime, layered conformance profiles, negative security tests, fuzz and abuse regression, and release-gate checks.
Evidence and scope
This arXiv preprint evaluates externally observable protocol invariants against a pinned minimal reference runtime. It does not claim production readiness. Several stronger controls remained partial or future work at that release, and production deployments require durable governance state, distributed admission control, and hardened domain adapters.
Original research
Accentrust research overview · Read the preprint · SSRN record · Research code
The date above belongs to the original research record. The publication date shown by this blog records when this overview was added to Figena. Research claims retain the scope and assumptions of the linked paper; listing the work here does not establish production availability.
