Everruns
Everruns Platform icon

Everruns Platform

The headless, durable agentic harness engine. Define agents and their tools, compose them into reusable harnesses, then ship them to real users.

← Back to products

GitHub: github.com/everruns/everruns · Docs: docs.everruns.com · License: MIT

What it is

Everruns is an open-source platform for building and deploying durable AI agents. You define agents and the tools they can use, compose them into reusable harnesses, and then ship those harnesses to real users through Slack, web chat, scheduled jobs, webhooks, A2A, MCP, or a plain HTTP API.

It sits between your application and the LLM providers, owning the agent loop — the cycle of reasoning (calling a model) and acting (running tools) — and making that loop durable. Every step and tool call is persisted, so a run survives process restarts, deploys, and infrastructure failures instead of starting over.

Architecture

The platform is built on a Rust and PostgreSQL durable execution engine — the Everruns Runtime.

  • Durable by default — PostgreSQL stores sessions and events. Workflows resume from where they stopped after a crash or restart.
  • Horizontally scalable — A control-plane / worker split communicates over gRPC. Workers carry no database credentials, so you can scale them out independently.
  • Real-time streaming — Server-Sent Events deliver live updates, backed by NATS JetStream in production or an in-memory broadcast in development.

Ship agents anywhere

The same agent definition can be exposed through multiple surfaces without rewriting it:

  • Slack and web chat for conversational use
  • Scheduled jobs and webhooks for automation
  • A2A and MCP for agent-to-agent and tool interoperability
  • A plain HTTP API for everything else

SDKs

Official client libraries provide a consistent, typed, async API across Rust, Python, and TypeScript. See the SDK documentation.