Share your experiences (with a model)
Summarized my recent addition to development workflows that greatly improved virtually everything in my setup.
korshakov.com/posts/share-yo…
Steve Korshakov
@Ex3NDR
Chaotic good inventor. Applied AI at @bee__computer. Father of Happy Coder. Past: Openland (YC W18), Actor Messenger, early employee Telegram.
- Introducing Prime Agent: A self-improving RLM harness for coding and long-running autonomous tasks. Designed to be both token-efficient and expressive through programmatic tool calling, context as a variable, multi-agent messaging, and a self-modifiable harness state.
- Tried @dignifiedquire very nice work to replace webrtc-like p2p networking! Hopefully soon in @happy__eng agents would be able to communicate together.
- very nice, does anyone found a good usage for luna? it is barely agentic and barely smart, so i can’t make up a usage besides simple classifications and summaries - does anyone use such models for anything else?major price cuts today: *80% drop for GPT-5.6 Luna, now $0.20 per million input tokens and $1.20 per million output *20% drop for GPT-5.6 Terra, to $2/$12 *GPT-5.6 Sol gets Fast mode in the API, up to 2.5x the speed for 2x the price, same intelligence
