Documentation

Welcome to the HiveBoss documentation. Here you will find everything you need to set up your organisation, understand how the Hive Mind works, and connect your own tools and AI agents to it.

What is HiveBoss?

HiveBoss is a team collaboration platform — chat, channels, and live calls ("Waggle") — built around an organisation-wide knowledge graph called the Hive Mind.

As your team works, an ingestion pipeline automatically extracts and structures knowledge from conversations, calls, and imported files (GitHub, Gmail, ZIP, plain text) into the Hive Mind — deduplicating and linking related information so it becomes a persistent, queryable memory for your organisation.

On top of that memory, you can configure AI agents (Bees) scoped to departments or roles. Bees can read and write the Hive Mind, run workflows, and answer questions in chat — consuming a metered usage currency called Honey, tracked per organisation.

What you can do

• Chat and collaborate — direct messages, group conversations, and channels
• Hold live calls — voice/video sessions ("Waggle") that produce recaps: decisions, action items, and open questions
• Build a Hive Mind — a knowledge graph that grows automatically from chat, files, and calls
• Configure Bees — department-scoped AI agents that can search, read, and write your Hive Mind
• Automate with workflows — multi-step automations that Bees can run
• Track usage — a Honey ledger meters AI/agent usage costs per organisation
• Connect external tools — over MCP, any MCP-compatible client (Claude, Cursor, Claude Code) can search and update your Hive Mind directly

Where to go next

• Getting Started — a tour of how the pieces (chat, Hive Mind, Bees, Honey) fit together
• MCP Server — connect an external AI agent to your Hive Mind over MCP