daydreamsai/daydreams

Daydreams is a set of tools for building agents for commerce

What it solves

Daydreams addresses the common pitfalls of AI agent development, such as context switching breaking functionality, nightmare state management, and lack of persistent memory across sessions. It prevents agents from becoming a "tangled mess of prompts and logic" as they scale in complexity.

How it works

The framework uses a Composable Context Architecture. Instead of a monolithic design, it creates isolated, stateful workspaces (contexts) that can be combined using a .use() method. This allows developers to build complex behaviors by composing modular components.

Key architectural components include:

  • Context System: Isolated workspaces for different conversation types with their own schemas and instructions.
  • Memory System: A dual-tier storage system consisting of Working Memory (temporary execution state) and Context Memory (persistent data).
  • Action System: Type-safe functions with schema validation that can access and modify context memory.
  • Dreams Router: A universal AI gateway providing a single API to access multiple models (OpenAI, Anthropic, Google, etc.) with built-in USDC micropayments.
  • MCP Integration: Native support for the Model Context Protocol, allowing agents to connect to external tools like databases, file systems, and web services.

Who it’s for

TypeScript developers building production-ready AI agents that require persistent memory, complex state management, and the ability to integrate with various LLM providers and external tools.

Highlights

  • Composable Contexts: Build complex agents by combining reusable, isolated contexts.
  • Native MCP Support: Instant access to a wide ecosystem of external tools and services.
  • Persistent Memory: True stateful agents that remember information across restarts.
  • Universal AI Gateway: Access multiple LLM providers through one API with pay-per-use micropayments.
  • TypeScript-first: Full type safety and excellent developer experience.
  • Universal Runtime: Compatible with Node.js, browsers, Deno, Bun, and edge functions.

Related

  • Project
    joeynyc/hermes-skins

    Hermes Skins is a collection of YAML-based visual themes for the Hermes CLI agent, letting users customize colors, spinner animations, branding, and ASCII art without altering the agent's behavior. It includes over a dozen curated skins and supports creating your own by dropping a YAML file into the skins directory.

  • Project
    MaxGfeller/open-harness

    A framework based on Vercel's AI SDK for building general-purpose AI agents with built-in support for tool use, session management, and composable middleware.

  • Project
    RekklesNA/ProxmoxMCP-Plus

    ProxmoxMCP-Plus is a bridge that lets AI agents (via MCP) and standard HTTP tools (via OpenAPI) safely operate a Proxmox VE virtualization server, covering VM/LXC lifecycle, snapshots, backups, ISOs, container commands, and persistent job tracking. It emphasizes security with API tokens, command policies, approval gates, and TLS validation, and supports local stdio, Docker-based HTTP, and an opt-in sandboxed Code Mode. An optional OAuth browser-gate flow backed by PostgreSQL handles remote client authentication.

  • Project
    hydro13/tandem-browser

    A local-first Electron browser that integrates AI agents directly into the runtime, allowing them to automate real SaaS platforms and interact with authenticated web sessions.