Why HiveLang Exists
If you’ve built agents before, you know the pain of managing scattered prompts, tool definitions, and deployment configs across multiple files. HiveLang consolidates everything into one readable format. HiveLang defines:- Agent identity and personality
- Clear instructions and rules
- Tool capabilities and workflows
- Event hooks and orchestration
- Multi-agent Swarms
Quick Example (v5)
Here’s a complete agent in HiveLang v5:Core Concepts
1. Bot Block
Every HiveLang file defines agents via abot (or agent) block:
2. Instructions
Instructions define behavior using curly braces:The v5 parser natively understands curly braces for multi-line instructions, eliminating the need for quotes.
3. Capabilities
Capabilities give agents access to integrations:4. Event Hooks
Define pre- and post-processing:5. Multi-Agent Swarms
Orchestrate complex workflows withswarm:
HiveLang vs Traditional Approaches
Compared to Raw Prompts
Python:HiveLang:
Compared to LangChain
LangChain is code-first and gives engineering teams deep control over agent architecture. HiveLang is spec-first: it keeps the agent’s identity, tools, instructions, schedules, memory behavior, and orchestration in one readable definition.Use LangChain when you want to build the system in Python or JavaScript. Use HiveLang when you want the agent definition to be portable across Bothive’s builder, runtime, review, deployment, and observability surfaces.