neuro-san and neuro-san-studio
The neuro-san library provides the core agent framework while neuro-san-studio serves as its interactive development and experimentation environment, making them complements designed to be used together.
About neuro-san
cognizant-ai-lab/neuro-san
Neuro AI System of Agent Networks
Decomposes complex LLM tasks into specialized multi-agent networks defined entirely in HOCON configuration files, enabling non-programmers to orchestrate agent workflows. Agents integrate deterministic CodedTools (LangChain or custom) for web APIs, private data handling, and computation that LLMs can't perform alone. Provides provider-agnostic LLM selection per-agent, private data channels, distributed agent webs, MCP protocol support, and observability integrations with LangSmith and Phoenix.
About neuro-san-studio
cognizant-ai-lab/neuro-san-studio
A playground for neuro-san
Provides declarative HOCON-based configuration for designing adaptive multi-agent networks with autonomous task delegation via the AAOSA protocol, eliminating extensive coding. Integrates seamlessly with multiple LLM providers (OpenAI, Anthropic, Azure, Ollama) and external ecosystems (CrewAI, LangChain, MCP), while offering specialized features like Sly-Data for secure sensitive data handling and a meta-agent for auto-generating network configurations. Studio includes ready-to-run examples, tutorials, and detailed tracing/logging infrastructure for rapid prototyping across finance, insurance, and enterprise domains.
Scores updated daily from GitHub, PyPI, and npm data. How scores work