Blades

One-Minute Overview#

Blades is a multimodal AI Agent framework for the Go language, providing a flexible and efficient AI solution through unified interfaces and pluggable components. It allows developers to easily build intelligent applications supporting multi-turn conversations, chain-of-thought reasoning, and structured output, making it ideal for integrating AI capabilities into Go projects.

Core Value: Offers a Go-native development experience for AI Agents with flexible extensibility and multi-model integration.

Quick Start#

Installation Difficulty: Medium - Requires Go environment and basic understanding of LLM/AI concepts

go get github.com/go-kratos/blades

Is this suitable for my needs?

• ✅ Developing Go-based AI applications: Blades provides Go-native experience that aligns with Go programming conventions
• ✅ Need multi-turn conversation memory: Built-in Memory component supports both short-term and long-term memory capabilities
• ✅ Need complex workflows: Chain components can be used to build complex multi-step reasoning processes
• ❌ Require simplest AI integration: The framework is feature-rich and might be overly complex for minimal use cases
• ❌ Not familiar with Go: The framework is completely Go-based and requires Go programming knowledge

Core Capabilities#

1. Agent - Task Executor#

The core execution unit in the Blades framework, capable of calling models and tools to complete tasks. Actual Value: Developers can define AI Agents through concise code declarations, enabling rapid requirement delivery and making complex logic clear and manageable.

2. ModelProvider - Multi-Model Unified Interface#

An abstraction layer for interacting with LLMs, allowing easy switching and integration of different language model services. Actual Value: Switch between different underlying AI models without modifying core code, avoiding vendor lock-in and improving system flexibility.

3. Tool - Capability Extension#

External functions or services that enable Agents to perform specific actions or obtain external information. Actual Value: Lets AI Agents interact with the real world through API calls, database queries, file system access, and other functions.

4. Memory - Conversation Memory System#

Provides short-term or long-term memory capabilities for Agents, ensuring contextual coherence across multi-turn conversations. Actual Value: Enables AI Agents to remember previous conversations, providing more coherent and personalized interaction experiences.

5. Middleware - Cross-Cutting Concerns#

A powerful mechanism for implementing cross-cutting concerns like logging, monitoring, authentication, and rate limiting. Actual Value: Add additional functionality to AI Agents through middleware chains without modifying core business logic.

6. Chain - Complex Workflow Construction#

Connects multiple Agents or Chains to form complex multi-step reasoning and workflows. Actual Value: Chain multiple intelligent components to implement complex decision-making processes and data processing flows.

Technology Stack & Integration#

Development Language: Go Key Dependencies: Standard Go libraries, OpenAI SDK for examples Integration Method: Library/API

Maintenance Status#

• Development Activity: Actively maintained, in early stages with continuous rapid iteration
• Recent Updates: Frequent commits, multiple times per week, seeking community feedback
• Community Response: Active GitHub project welcoming contributions and feedback

Documentation & Learning Resources#

• Documentation Quality: Comprehensive, including architecture design, core concepts, and usage examples
• Official Documentation: https://github.com/go-kratos/blades/tree/main/docs
• Example Code: https://github.com/go-kratos/blades/tree/main/examples