Stop AI agents from doing things they shouldn't. The action authorization layer — pre-execution, deterministic, policies included. Drop in. Day one.

The authorization kernel for AI agents. Block, allow, or escalate agent tool calls with YAML rules — deterministic-first, LLM fallback.

Authorization, delegation, provenance, and verifiable-audit engine for AI agents. MCP adapter published.

Draco-Tech is the identity and authorization layer that lets autonomous agents operate with zero static credentials and full policy adherence.

EVVM user validators using the ERC-8004 Identity Registry for agent authorization and whitelist management.

An AI agent whose every action is authorized by a revocable credential and sealed into a tamper-evident log — and that reconciles actions taken offline on a credential revoked while it couldn't phone home. With a forensics view that shows the recomputed bytes, not just a verdict. A reference implementation on provenance-core.

pgauthz brings Google Zanzibar-style (ReBAC) authorization directly into PostgreSQL for multiple Agents workflows

Open-source runtime authorization and evidence controls for AI agents.

HARP: Healthcare Agent Registration and Connection Profile. An open MIT standard for AI agents to register, accept a BAA by machine, and act under least-privilege purpose-bound access in healthcare revenue operations.

Agent action authorization and signed audit control plane for Zero Trust infrastructure

Reference implementation of policy-embedded credential authorization for AI agents (arXiv:2605.11487). Issue signed credentials with machine-evaluable constraints, verify at runtime, produce signed audit decisions.

Fail-closed Python runtime for AI agent authorization, seed-mediated authority migration, canonical JSON ledgers, and staged declared autonomy for RLHF-shaped agents.

AMI — Agent Mobile Identity Protocol. Let an AI agent request, contract (real KYC), activate and operate a real mobile identity — number, voice, SMS, data — over MCP (28 tools) and a REST API. Identity + fine-grained permissions for agents, on top of MCP and A2A.

OAuth-style delegated authorization for AI agents and MCP servers — scoped, revocable, short-lived tokens.

A reference implementation of a stateful, HITL-aware, parameter-bound bridge between A2A and MCP.

A vendor-neutral protocol for recording why, by whom, under which policies, scopes, risks, and human-review conditions an AI agent action was authorized, restricted, deferred, or denied.

A deterministic agent registry — verify an AI agent is real, was delegated by a verified human, and that the delegation is still valid. Verify-first, stateless, source-available.

agents.txt — an action-policy format letting sites declare what autonomous agents may do, with verifiable receipts

Least-privilege authorization for enterprise AI agents. Orchestrator delegates scoped, time-limited, revocable credentials to sub-agents via Terminal 3 Agent Auth SDK. Violations rejected before execution. Immutable signed audit trail. Built for banks, governments, and institutions.

Open authority and audit protocol for AI agents: scoped delegation, human approval, portable authorization, signed records for bounded audit, and agentic commerce.