dbreunig.com
Why longer-running agents still need fresh context.
Drew Breunig’s October 24 talk explains how context poisoning, distraction and inertia derail agents. Slides, examples and DSPy code show ways to select tools, prune history and reset context while preserving work. Practical reading for indie hackers building agents or debugging their own coding workflows.
AI regulatory intelligence platform 🌱 Sponsored
Anthropic's model for long-running agents and coding.
Google previews its next model for complex, long-running work.
Enhancing coding agents with advanced context techniques.
Thoughts on long-context AI models by Steven Johnson
Standardizing context for AI applications
A brand new look, a fresh way to browse, a new start
Why usage-based services need limits that actually stop spending.
Chat with strangers while watching viral content live.
Search listening tool for market,customer & content research
Transform text into stunning images — free, fast, and customizabl 🌱 Sponsored
A multiplayer environment for coding with agents.
Running a big model on a small laptop.
Fast, CPU-based code search library for agents.
AI coding dream team of agents for VS Code.
No comments yet. Have you tried it?