astro turf

For new followers: - I'm a long-time investor and builder in this space. - Founding Contributor of @LootRealms. - Co-founder of @ohayo_dojo. - Builder with the kings at @cartridge_gg. - @Starknet class of '21. - Founder and Game Director of @RealmsEternum. - Founder of @daydreamsagents My prime purpose for the past three years has been to build onchain infrastructure to enable the next generation of onchain experiences. This is done @Starknet as it is the superior VM for building complex applications—this will become clear soon enough. I work up and down the entire stack, from low-level indexing and contracts to GUI design. Nothing is out of scope. I have been pushing on agents for two years, mostly using existing frameworks like @langchain, until I came across @ElizaOS_ai in October. As I focused on building agents for @RealmsEternum, it became clear that agents playing games require infinite paths to achieve goals. Thus, it's not scalable to hardcode functions—agents need to have total fluidity to take any action or call anything the game requires in any order. And ironically onchain infra is perfect for agent playgrounds because of its open nature. This exploration led me to create @daydreamsagents, which focuses on the hardest problems of agents: long time-horizon goals using Hierarchical task networks (HTN). Daydreams agents don't require custom code—they work entirely based on 'sleeves'—which are just markdown files that explain how the agent can interact with the service (API docs, game guides, etc.) My thesis is simple. By focusing on the hardest problem (games), the design of the library will naturally lean towards an optimal structure for any problem an agent could face. We are early in this path and iterating with speed. If you are an onchain app developer or game builder—DM me, I want to know the architecture of your game so we can build sleeves together.

My original thesis for founding @daydreamsagents was to tackle the hardest problems, thereby forcing the evolution of agent architecture and the creation of enriched logical data to train the next generation of agents. It’s fitting that the other major open-source project I founded with some other chads is named “dojo” (@ohayo_dojo), whose main purpose was to accelerate the creation of onchain games. As the world begins to understand that we no longer need raw data but rather logical pathways and solutions to train new models, it will become clear that games are the way to do this. Ironically, onchain games are the perfect fit due to their inherent openness and lower stakes. It’s preordained at this point, with even the influential @karpathy openly discussing games right now. There are already a bunch of games hacking on @daydreamsagents to achieve this goal—but we need to go further and create an accelerated environment for onchain games. If you have a game and want to push hard on this, please reach out. The tighter we make the development cycle of agents building logical pathways in games, the faster our models accelerate—and the sooner we achieve AGI.

v0.0.4 @daydreamsagents out - 4 new Contributors - Greater state abstraction - Cleanups and iterations - @HyperliquidX example - New websocket streaming IO - Telegram example This is a fairly minor patch as we push on to a 0.1.0 which will introduce a more stable architecture Our north star is the simplest most poweful implementation of a stateful chain-of-thought Time to jump in dreamer

One thing that drove me to build @daydreamsagents was the lack of advanced chain of thought in agent frameworks to accomplish hard stateful goals. Everyone is seeing power of chain of thought at the model level with deepseek now - which uses the technique deeply during inference. What daydreams does is a layer above. It allows chaining - chain of thought at the agent level to accomplish goals over long periods of time. So if you use deepseek with daydreams it’s two layers of thought with the ability to work across potentially any other service ok the internet(defai, games, equities) We only just scratching the surface here If you are interested in these - let’s collaborate on the next generation of agents

dreaming @daydreamsagents