4 repository-uri
Control mechanisms for autonomous agents using persistent logs and approval boundaries to ensure reliable execution.
Distinct from Durable Execution Persistence: Distinct from Durable Execution Persistence: focuses on the behavioral control and approval gates of an agent loop rather than just crash recovery.
Explore 4 awesome GitHub repositories matching software engineering & architecture · Agent Execution Loops. Refine with filters or upvote what's useful.
This project is a comprehensive framework for the orchestration, evaluation, and context management of large language model agents. It provides a set of architectural patterns and standards for designing agent interactions, integrating external tools, and establishing memory architectures to persist knowledge across sessions. The system focuses on optimizing the limited memory of language models through token-aware context compression and filesystem-based context offloading. It incorporates secure execution environments using sandboxed virtual machines and isolated containers to safely run ba
Implements durable execution loops with locked metrics and human-in-the-loop approval boundaries.
ZenML is an extensible machine learning orchestration framework designed to manage the end-to-end lifecycle of data pipelines and AI agent workflows. It functions as a durable orchestrator that executes machine learning tasks as directed acyclic graphs, ensuring that every step is containerized for consistent performance across local, cloud, and hybrid infrastructure. By decoupling pipeline code from underlying compute and storage backends, the platform allows developers to define infrastructure-agnostic stacks that remain portable across diverse environments. The project distinguishes itself
Refactors existing agentic code from various SDKs into a durable execution model.
ZenML is an orchestration platform designed for building, deploying, and monitoring reproducible machine learning pipelines and agentic workflows. It provides a unified framework that manages the entire lifecycle of machine learning assets, from data processing and model training to the deployment of persistent inference services. By decoupling pipeline logic from underlying compute and storage, the platform enables teams to transition workflows seamlessly from local development environments to production-grade cloud infrastructure. The platform distinguishes itself through a service-oriented
Wraps agent logic in durable, checkpointed flows to enable stateful resumption, replay, and auditability of complex interactions.
OpenSquilla este un framework de orchestrare pentru agenți LLM, conceput pentru a coordona fluxuri de lucru AI în mai mulți pași și execuția de instrumente folosind grafuri aciclice direcționate. Acesta funcționează ca un sistem centralizat pentru gestionarea pachetelor de competențe specializate și executarea secvențelor complexe de raționament. Proiectul se distinge printr-un gateway de rutare care direcționează sarcinile către diferiți furnizori AI în funcție de complexitate, cost și performanță. Utilizează un sistem de memorie AI pe mai multe niveluri care organizează cunoștințele de lucru, episodice și semantice folosind embedding-uri locale și SQLite, alături de un sandbox de execuție securizat care izolează codul generat de agenți prin profiluri de permisiuni bazate pe risc. Platforma acoperă o gamă largă de capabilități, inclusiv implementarea pe mai multe canale către platforme web și de mesagerie, programarea automată a sarcinilor prin cron și un bridge Model Context Protocol pentru conectarea la instrumente externe. De asemenea, oferă instrumente cuprinzătoare de monitorizare și observabilitate pentru urmărirea costurilor per token, auditarea deciziilor runtime și gestionarea unui catalog de competențe reutilizabile. Sistemul include utilitare CLI pentru inițializarea spațiului de lucru și gestionarea ciclului de viață al competențelor.
Provides tools to import configuration and state from legacy agent frameworks into the current orchestration system.