# NimbleBrain

> NimbleBrain is an AI advisory and product company. We embed senior engineers who ship production AI agent systems in weeks, then transfer full ownership to the client. No lock-in, no ongoing dependency.

## What we do
- AI advisory and delivery: production agent systems built on an open-source stack, handed off with full ownership
- Business-as-Code: operations encoded as schemas + skills + context so agents can read and execute them
- Deep Agent Method: our methodology for organizing business knowledge for AI

## Products
- [Upjack](https://upjack.dev): declarative AI application framework — define apps as JSON schemas + natural-language skills
- [mpak](https://mpak.dev): open-source MCP server registry with security scanning and L1–L4 trust scoring
- [Synapse](https://nimblebrain.ai/products/): protocol-native UI for AI agents, generated from MCP tool schemas
- [Bassethound](https://bassethound.ai): remote MCP for deep, correlated domain intelligence
- [MCP Trust Framework](https://mpaktrust.org): security standard for MCP tool distribution

## MCP endpoints
- Bassethound remote MCP (Streamable HTTP): https://mcp.bassethound.ai/mcp
- Discover more MCP servers: https://mpak.dev

## Key pages
- Deep Agent Method: https://nimblebrain.ai/method/
- Production AI Playbook: https://nimblebrain.ai/guides/production-ai-playbook/
- Business-as-Code Guide: https://nimblebrain.ai/guides/business-as-code-guide/
- Products: https://nimblebrain.ai/products/
- Technology: https://nimblebrain.ai/technology/
- Industries: https://nimblebrain.ai/verticals/

## Company
- Founded 2024 · Honolulu, HI
- CEO & Co-Founder: Mat Goldsborough
- Contact: hello@nimblebrain.ai
- Documentation: https://docs.nimblebrain.ai
- Security contact: https://nimblebrain.ai/.well-known/security.txt

## Optional
- Full text for LLMs: https://nimblebrain.ai/llms-full.txt