execution-boundary

Fail-closed execution firewall for AI agents: quarantine MCP tools, proxy OpenAI-compatible requests, emit signed receipts, and verify EvidencePacks offline.

Execution boundary for autonomous systems. Deterministic runtime enforcement of executable constraints before state-changing actions occur.

Deterministic pre-execution gate for AI agents (fail-closed, YAML policy)

Workspace runtime and execution boundary for AI agents

Deterministic execution boundary for AI agents. IFC enforcement at the sink. 5 frameworks. 50 attack vectors. Apache 2.0.

Deterministic three-state stop controller. Constraint-first design. Fail-closed.

Core execution boundary engine (decision + proof + ledger)

Transport-layer application profile for Execution Boundary Core Spec.

AI Execution Boundary Standard (AEBS) - Pre-incident reference for execution governance

Execution boundary for GitHub pull requests that interprets repository mutations before CI enforcement.

Observability profile for Execution Boundary systems — OTel collector topology, semantic conventions, dashboards, alerts

Small, deterministic control primitives for software systems. Testable, auditable bricks.

A Runtime Control Model for Autonomous Systems — Working Paper V2.0

Admissibility Rotation Corridor — closed constitutional runtime with visible interrupt geometry

Source-visible bounded proof/witness stack for execution integrity, replay-legitimacy separation, and consequence-bearing runtime governance.

Execution-boundary gate interface. Contamination cases. Adversarial suite. Admissibility at execution time.

Deterministic admissibility layer for interpretation proposals before verdict and execution.

Public-safe PoC for admissibility-first execution governance in probabilistic and multi-agent AI systems.

Minimal proof of execution-bound admissibility: no present-state proof at bind, no execution.

Entry point and architectural map for execution boundary research.