aden-hive/hive
Hive
Hive is an artificial intelligence workflow automation engine and development platform designed for building and deploying autonomous agents. It provides a framework for orchestrating complex, multi-step business processes by coordinating tasks across multiple specialized agents using directed graph structures.
The platform distinguishes itself through a focus on production-grade reliability and state management. It maintains persistent execution context and conversation history on disk, enabling crash recovery and continuity for long-running automated sessions. Furthermore, it incorporates a multi-level evaluation pipeline that validates agent outputs through a combination of deterministic rules, semantic quality assessments, and human oversight.
The system includes a universal model abstraction layer that allows developers to interface with diverse local or hosted language models. It also features operational policy enforcement, providing real-time metrics, budget controls, and audit trails to monitor workloads. To manage memory and performance, the engine optimizes context windows by truncating large data into compact references that agents can retrieve on demand.
Features
- Awesome List - A community-curated directory that catalogs and links out to other open-source projects, rather than a standalone tool you run yourself.
- Agentic LLM Frameworks - Serves as a platform for building and deploying intelligent agents that integrate with external tools and diverse language models.
- AI Workflow Automation - Provides an engine for managing multi-step automated tasks with iterative feedback, human oversight, and operational policy enforcement.
- Multi-Agent Orchestration - Orchestrates complex business processes by coordinating tasks across multiple specialized agents using directed graph structures.
- Multi-Agent Coordination Systems - Orchestrates parallel task execution across multiple specialized agents using graph-based structures.
- Agent State Persistence - Maintains persistent execution context and conversation history on disk to ensure crash recovery and continuity for long-running automated sessions.
- Multi-Agent Orchestration Frameworks - Coordinates complex workflows across multiple autonomous agents using graph-based task execution and persistent state management.
- AI Agent Tool Integrations - Connects autonomous agents to external business systems and internal APIs for real-world task execution.
- Model Abstractions - Provides a unified interface to interact with diverse local or hosted language models.
- Operational Policy Enforcement - Provides real-time metrics, budget controls, and audit trails to monitor and enforce operational policies for automated workloads.
- Context Compaction Engines - Optimizes context windows by truncating large data into compact references for on-demand retrieval.
- Directed Acyclic Graph Engines - Orchestrates complex workflows by routing tasks through a sequence of nodes using directed acyclic graphs.
- Automated Agent Quality Assurance - Validates agent performance through a multi-level pipeline of deterministic rules, semantic assessment, and human oversight.
- Agent Tool Integrations - Connects autonomous agents to external business systems and APIs using standardized protocols.
- Conversation State Persistence - Serializes execution context and conversation history to disk to ensure crash recovery and continuity.
- External Tool Integrations - Integrates agents with external business systems and APIs to perform real-world actions.
- Multi-Agent Output Evaluation - Validates agent outputs using a multi-level pipeline of deterministic rules, semantic assessment, and human oversight.
- Local Language Model Execution - Supports diverse local or hosted language models through a universal interface.
- Agent Frameworks - Framework for building goal-driven, self-improving autonomous agents.
- AI Agent Frameworks - Outcome-driven agent development framework.
- Autonomous AI Agents - Multi-agent framework with auto-generated graphs and evolution loops.
- Agent State Persistence - Persists agent execution context to enable crash recovery and session continuity.
- Persistent State Management - Maintains execution state and conversation logs across disk and memory for long-running workflows.
- Execution Graphs - Executes work through directed graphs where nodes maintain independent state and instructions.
- Context Truncators - Optimizes context windows by truncating data into compact references for on-demand retrieval.
- Memory Reflection Automations - Injects feedback from automated judges and human reviewers back into conversation history for iterative refinement.
- Configuration and Policy Enforcement - Enforces operational policies through real-time metrics, budget controls, and audit trails for production workloads.
- State Persistence Layers - Maintains persistent execution state across long-running automated sessions to enable crash recovery.
Star history
How this analysis was created: This summary and feature list are AI-generated from collected project material and can contain mistakes. Stars, license and language are imported from GitHub. Inclusion does not mean that we have tested or audited this project. Check the source documentation for any feature you depend on. Learn more on our About page.
