AgentRun launches open-source control plane for AI agents
AgentRun is an Apache-2.0 open-source platform that adds self-improvement, durable execution and a central control plane around existing AI-agent harnesses. It supports pi, Claude Code, Codex and OpenCode, with persistent learning, run recovery and tracking, approvals, spending caps, triggers, and environments such as sandboxes, desktops, phones and browsers. Parcha Labs holds the copyright, and Grep.ai built the platform. The vendor says agents become more fault-tolerant and use fewer tokens over repeated runs. It also reports that its own code-reviewer setup costs 18 cents per review, but the source provides no independent measurement of that figure.
AgentRun offers an open-source way to add persistent learning, recovery and operational controls across several AI-agent harnesses. The reported efficiency and cost benefits have not been independently validated.
Run smarter agents. Use fewer tokens.
Our code review bill went up 36x this year. Now a review costs us 18 cents. See how ↓
Quickstart · Case studies · Tour · Templates · Deploy · Docs
AgentRun is an open-source platform that supercharges your agents. It adds self-improvement, durable execution and a control plane on top of your favorite agent harness. Your agents become smarter, more fault tolerant and use fewer tokens with every run.

An agent's Learning tab: what it kept, and what changed since. Demo data from the console's recorded fixtures.
#AgentRun#Parcha Labs#Grep.ai#pi#Claude Code#Codex#OpenCode#Apache-2.0