Breakpoints for Swift Testing - precise control over side effects and fully observable state transitions in asynchronous functions
-
Updated
Nov 16, 2025 - Swift
Breakpoints for Swift Testing - precise control over side effects and fully observable state transitions in asynchronous functions
Execution control layer for AI agents — prevents duplicate or incorrect real-world actions under retries, uncertainty, and stale context.
The open protocol for persistent and verifiable governance of autonomous AI agents. Four-layer substrate: Persistent Context, Dynamic Governance, Deterministic Admissibility, Auditable Ledger. 0.07ms enforcement. Zero LLM inference.
Radium corpus — Execution Governance for Complex Automated Systems
Vienna OS — The execution control layer for autonomous AI systems. Cryptographic warrants, risk tiering, policy enforcement, and immutable audit trails.
Sample project showcasing capabilities of the Probing package
Legacy deterministic Event-driven trading decision engine for shared Backtesting and Live semantics.
how to use the rule-flow to orchestrate DRL rules
Local-first execution-control reference runtime for AI agent builders.
A deep instrumentation layer built over the sval JS-in-JS interpreter to monitor functions as they execute.
Simple modifications to Coffeescript source to enable pausing of a program, and resuming based on simple conditions e.g. time period
Deterministic, tick bound runtime integrity verification that blocks execution under drift or tamper before any sensitive operation.
AI Decision Validator (AI Execution Firewall) Validates AI-generated decisions before execution using deterministic tests, fail-closed validation, and cryptographic audit.
Fail-closed enforcement layer that converts deterministic ACCEPT/REJECT decisions into guaranteed execution control with a public verification URL.
A neutral execution-control seam for agent tool calls: allow / block / indeterminate, only allow proceeds.
.NET runtime of the Spine Lite kernel. Policy-gated execution + receipts. Built on Semantic Kernel. MIT.
PFC separates the time authority from the signer so a compromised enforcement point can't silently backdate its own approvals — and makes replay and reordering detectable through monotonic, hash-chained receipts.
Provable execution membrane for high-stakes AI actions, with offline-verifiable governance bundles.
Python runtime of the Spine Lite kernel. Policy-gated execution + receipts for AI agents. MIT.
PowerShell research tool showing EXE execution control bypasses in legacy Windows systems (up to Windows 10) for authorized testing.
Add a description, image, and links to the execution-control topic page so that developers can more easily learn about it.
To associate your repository with the execution-control topic, visit your repo's landing page and select "manage topics."