Reference
Technical Architecture
Written by Hive Intelligence, Product docsLast updated
Overview
The Hive server is structured around a provider-backed tool catalog plus Hive-native stateful tools, with multiple access surfaces layered on top.
Main Layers
Wrapped provider catalog
This is the live provider-backed tool surface used by the server runtime.
- 351 wrapped provider tools
- provider-backed handlers
- shared schema and metadata normalization
Hive-native stateful layer
This layer adds persistence on top of provider calls.
- 18 Hive-native tools
- monitor, alert, observation, memory, report, and B2B subject handlers
- subject-scoped state for partner integrations
Root MCP endpoint
The root /mcp endpoint is the recommended AI-facing surface.
- compact discovery-oriented
tools/list - resource-based discovery through
hive://tools,hive://categories,hive://providers,hive://toolsets,hive://task-canaries,hive://skills, andhive://status - exact tool execution through
invoke_api_endpoint
Category-scoped MCP endpoints
These are narrower surfaces built from toolRegistry.ts.
- 10 scoped endpoints
- 351 categorized provider tools
- direct
tools/listresults by category
REST API
The REST layer exposes:
GET /api/v1/toolsfor the full 369-tool callable catalogPOST /api/v1/executefor tool executionGET /api/v1/usagefor plan, credit wallet, and key summaryGET|POST|DELETE /api/v1/keysfor dashboard key managementGET /api/openapi.jsonand/docsfor OpenAPI documentation/status,/api/health/*,/live,/ready, and/startupfor health and canaries
Provider Routing
Requests are routed to provider-specific handlers based on tool metadata.
Current live provider inventory:
- CoinGecko
- Alchemy
- DeFiLlama
- GoPlus
- Codex
- CCXT
- Helius
- Tenderly
- Moralis
Discovery Model
There are two important discovery patterns:
- Root discovery for task routing, schema lookup, and
invoke_api_endpoint - Category-scoped discovery for smaller, direct tool lists
This distinction is the main reason the docs now separate Tools Reference from Live Catalog.