AgentRun open-sources self-improving agent platform that cuts repeated token spend
LinusEkenstam · x · 2026-11-11
Developer @aj announced AgentRun, an open-source self-improving agent platform and control plane. The core insight: agents figure things out repeatedly — rediscovering the same APIs and rereading the same files on every run, wasting tokens on answers they already found. Labs aren't incentivized to fix this, so the team built it themselves.
- Agents accumulate experience as they run, improving themselves and consuming fewer tokens over time.
- Works with existing Claude, Codex, and Pi agents as an external control plane.
- The launch post explains the motivation and mechanism for turning repeated agent exploration into reusable memory.
Related event: AgentRun Open-Sources Self-Improving Agent Platform to Cut Token Waste(3 posts)→
More from coding & agent
- Supabase lands in Expo EAS CLI: one command wires a backend into every environment — dshukertjr · 2026-11-11
- What would you rewrite with AI? Orosz sparks a list of cheaper rebuilds of existing services — _philschmid · 2026-11-11
- Ex-Meta researcher predicts a hard ceiling of 150-200B tokens for the SWE market — DimitrisPapail · 2026-11-11
- ComfyUI Challenge Winner ComfyArt: Open-Source Fully Local AI Design Canvas — bfl_ai · 2026-11-11
- Letting an agent self-improve its own harness daily: resumed runs 7→77, memory noise 94%→0 — AccBalanced · 2026-11-11
- One Photo of a Toddler's Playroom Becomes a Racing Game, Then the Whole House — every · 2026-11-11