polos-dev/polos
The open-source runtime for AI agents. Sandboxed execution with built-in tools, human-in-the-loop approvals, Slack integration, and durable workflows with automatic retries and prompt caching. You write the agent. Polos handles the infrastructure.
# Technical Summary Polos implements durable agent execution through event sourcing—every tool call and API response is logged, allowing workflows to resume from failure without re-executing completed steps or re-paying for LLM calls. The architecture separates a Rust-based orchestrator (managing state, retries, scheduling, and Postgres persistence) from horizontally-scalable Python/TypeScript workers that run agents via SDK, connecting over a persistent transport. It integrates framework-agnostic: agents built with CrewAI, LangGraph, or Mastra can adopt Polos's durability and observability without rewriting, while new agents use native Python or TypeScript—no DAGs or graph syntax required.
Stars
26
Forks
4
Language
TypeScript
License
Apache-2.0
Category
Last pushed
Feb 24, 2026
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/agents/polos-dev/polos"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Featured in
Higher-rated alternatives
agno-agi/agno
Build, run, manage agentic software at scale.
agentscope-ai/agentscope
Build and run agents you can see, understand and trust.
iflytek/astron-agent
Enterprise-grade, commercial-friendly agentic workflow platform for building next-generation SuperAgents.
adaline/gateway
The only fully local production-grade Super SDK that provides a simple, unified, and powerful...
agentuse/agentuse
🤖 AI agents on autopilot. Any model. Runs local, cron, CI/CD, or Docker.