The request asks what to use for agent development without naming a product, so these are the candidate frameworks and SDKs for building AI agents. AI SDK is a unified TypeScript toolkit for AI apps and agents with multi-model support 1, Genkit is Google's open-source framework for agentic apps in TypeScript, Go, Python and Dart 3, and Strands Agents is an open-source toolkit for production agents in Python and TypeScript 5. Which language or platform you build on, and whether you need voi
A question back: Which language or platform will the agent run on (TypeScript, Python, Go, Dart), and is it a text, voice or physical agent?
- 1
AI SDKai-sdk.devverified and signed
A unified TypeScript SDK for building AI apps and agents with streaming, fallbacks and multi-model support, with an MCP server and an ai-sdk skill for building AI features 111.
official:MCP server · @ai-sdk/mcpSDK · aiskill · ai-sdkskill · Skillsskill · building-components
for agents: eve · build-agents · ai-sdk
- 2
Genkitgenkit.devverified and signed
Google's open-source framework for building AI-powered apps and agents in TypeScript, Go, Python and Dart, with agent-building skills per language 313.
+2 community tools built for it
for agents: developing-genkit-dart · developing-genkit-go · developing-genkit-python
- 3
- 4
- 5
Sources
- 1AI SDKai-sdk.dev · read 2026-10-29
- 3Genkit | Google's open-source framework for AI apps & agentsgenkit.dev · read 2026-10-30
- 4Build voice, video, and physical AI | LiveKitlivekit.com · read 2026-10-29
- 5Strands Agents | The open source toolkit for production AI agentsstrandsagents.com · read 2026-10-29
- 9AgentsKit — Build agents without gluing many libraries together. | AgentsKitagentskit.io · read 2026-10-29
- 11eve: declares "eve" (the record we read and signed)unfragile.ai · read 2026-10-24
- 13developing-genkit-dart: declares "developing-genkit-dart" (the record we read and signed)unfragile.ai · read 2026-10-24
- 14livekit-agents: declares "livekit-agents" (the record we read and signed)unfragile.ai · read 2026-10-24
For agents: the 10 actions behind this answer
eve · eve verified and signed
Build durable backend AI agents with the eve framework. Use when creating, editing, or debugging an eve project — agent instructions, skills, tools, connections, channels, sandboxes, subagents, schedules, or evals.
microsoft-agent-framework · microsoft-agent-framework verified and signed
Create, update, refactor, explain, or review Microsoft Agent Framework solutions using shared guidance plus language-specific references for .NET and Python.
developing-genkit-dart · developing-genkit-dart verified and signed
Generates code and provides documentation for the Genkit Dart SDK. Use when the user asks to build AI agents in Dart, use Genkit flows, or integrate LLMs into Dart/Flutter applications.
livekit-agents · livekit-agents verified and signed
Build voice AI agents with LiveKit Cloud and the Agents SDK. Use when the user asks to "build a voice agent", "create a LiveKit agent", "add voice AI", "implement handoffs", "structure agent workflows", or is working with LiveKit Agents SDK. Provides opinionated guidance for the recommended path: Li
· Strands Agents website read, tools not yet checked
Strands is the open source toolkit for building production AI agents in Python and TypeScript: an agent harness and SDK, sandboxed shell, evals, and research labs.
· Symbolica website read, tools not yet checked
Build agentic applications with the Agentica SDK
· Robota website read, tools not yet checked
Build AI agents with the Robota TypeScript SDK, or use the coding CLI built from the same public components. Available in public beta.
agents · nemo-relay verified and signed
· AgentsKit website read, tools not yet checked
Build agents without gluing many libraries together. A composable TypeScript foundation for runtime, tools, memory, RAG, and chat interfaces.
did_init · afs found in code, not yet checked
Generate DID identity for developer, provider, or blocklet
Labels are the instruments' verdict and are never changed by ranking. Agents and robots get this same answer as JSON at /api/v1/find?q=…&answer=1 and through the MCP server.