2FastLabs/agent-squad

Project overview

2fastlabs/agent-squad is a developer framework for orchestrating multiple specialized AI agents, classifying user intent to route queries, and maintaining conversation context. It requires writing code to instantiate agents and typically depends on paid access to external LLM APIs.

Project type

AI Agent · Workflow

Use cases

Knowledge Q&A · Automation

Deployment

Refer to project documentation

License

Apache-2.0

Best for

• Developers and AI engineers building multi-agent workflows who need extensible components for routing and context management.

Key capabilities

• Analyzes user input to route queries to the most suitable specialized agent based on context and content.
• Maintains coherent conversations across different agents and sessions.
• Supports both streaming and non-streaming agent responses.
• Allows plugging in custom agents, classifiers, storage, and retrievers.
• Includes pre-built agents for Bedrock and Lex.
• A lead agent coordinates a team of specialized agents in parallel using an agent-as-tools architecture.
• Uses two LLMs (gatherer and isolated presenter) to ensure answers match fetched data exactly without drifting.
• Ships single, assistant-agnostic skill files providing API signatures and task recipes for AI assistants to write correct framework code.

Limitations and risks

• Requires coding knowledge to implement and use the framework.
• Running agents typically requires paid access to external LLM APIs.

Getting started

• Install the package via pip or npm, instantiate the orchestrator, and call route_request with a query. Setup difficulty is medium due to the need to write code to define agents and multiple runtime options.

Evidence and sources

• GitHub project description: Flexible and powerful framework for managing multiple AI agents and handling complex conversations
• README: Agent Squad routes each user query to the most suitable of your specialized agents and maintains conversation context across them. You get pre-built agents, classifiers, and stora…
• README: Python and TypeScript maintain feature parity and run anywhere — AWS Lambda, containers, your laptop. The new **Swift runtime** brings the same orchestration model to Apple platfo…
• README: npm install agent-squad
• README: 1. User input is analyzed by a **Classifier**. 2. The Classifier uses the agents' descriptions and the conversation history to select the best agent for the turn.

Reviewed at: 2026-07-17T00:56:01Z