runtime-policy

Trust infrastructure for autonomous AI agents: decision assurance, action authorization, runtime controls, and verifiable lifecycle evidence.

Local runtime security and policy enforcement for AI agents and LLM tool calls, with permissions, guardrails, rate limits, approvals, auditing, and an emergency kill switch.

Runtime policy guards that block repeated failing tool calls, oversized shell commands and network commands chained with local ones, composite batch tools (read_batch/edit_batch), SYSTEM.md replacement mode, soft convergence notices, and turn telemetry for DeepSeek Harness.

Enforcement gateway for AI agents. Nullrun official Python SDK.

Actra - control what runs before it runs, controls what actions are allowed before they execute. Evaluate policies across APIs, workflows and AI agents in real time.

GuardMesh — Portable policy checks for governed agent execution.

NullRun organisation profile — default community health files

Runtime execution control for AI agents. Policy enforcement, risk tracking, and three-strike shutdown, with a Python SDK for drop-in integration.

MCP server demonstrating server-side authorization with Stage0 runtime policy validation

Host-owned behavioral policy and runtime governance for AI agents: safe action gates, durable combined state authority, evidence, recovery, telemetry, and optional CSW/DP integration.

DeepSeek Harness plugin governance: installable runtime tool policy plus companion boot admission, supervised restart, and rollback.

Evidence-aware state-machine framework for research workflows with explicit claims, evidence, conflicts, runtime policy, checkpoints, provenance, claim verification, bounded transfer, and portable evidence handoff.