Our take
Activeregular commits and releases- Activity
- 90
- Community
- 30
- Issues
- 47
- Pull requests
- 77
+4 stars in the last 29 days
- Releases more often than 93% of the projects we track
- Merges more pull requests than 83% of the projects we track
AgentsKit is an actively maintained, broad JavaScript agent toolkit, but most contributions come from one person and most packages are still beta.
Good fit if
- You are building an agent in JavaScript or TypeScript and want runtime, tools, memory, and UI options in one ecosystem.
- You value provider and package substitution through shared contracts, and can adopt beta packages selectively.
- You can work with Node.js and want an MIT-licensed project with frequent recent commits and releases.
Look elsewhere if
- You only need one OpenAI streaming call, where the README itself recommends using the OpenAI SDK directly.
- You need Python, or need mature enterprise observability now; the project is JavaScript-first and describes its observability layer as young.
- Your team cannot accept reliance on a project where contributions are heavily concentrated in one contributor.
All health signals
| Last commit | 2026-10-13 (6 days ago) |
|---|---|
| Commits, last 90 days | 100+ |
| Releases, last 12 months | 100+ (latest @agentskit/vue@0.5.10, 2026-10-10) |
| Contributors | 6 (top contributor: 85% of commits) |
| Issues closed, last 90 days | 1+ |
| Pull requests merged, last 90 days | 72 (typically merged in 0 days) |
| Project age | 7 months |
Checked on 2026-10-20 with the GitHub API.
Project history
Daily GitHub snapshots since Oct 7, 2026. History covers stars and forks; health signals are not tracked historically.
Stars
38Forks
3Change since Nov 4, 2026: 0 stars·0 forks
Overview
AgentsKit is a TypeScript toolkit for building AI agents in JavaScript. It aims to replace one-off integrations between separate chat, runtime, tools, and memory libraries with composable packages connected by shared contracts. The core is designed to stay lightweight, while optional packages add autonomous execution, retrieval, observability, and other capabilities.
The toolkit supports React and terminal interfaces, alongside bindings for Vue, Svelte, Solid, Angular, and React Native. It is aimed at JavaScript developers who want to start with a single building block and expand toward a full agent stack without tying their application to one provider or UI framework.
Key Features
- A small, zero-runtime-dependency core with contracts for adapters, tools, memory, retrievers, skills, and runtime.
- Provider adapters for services and models including OpenAI, Anthropic, Gemini, Ollama, DeepSeek, and Grok.
- An autonomous runtime with tool use and multi-agent delegation.
- Packages for chat and vector memory, retrieval-augmented generation, and ready-made skills.
- UI options spanning React, terminal, Vue, Svelte, Solid, Angular, and React Native.
- Tool integrations, MCP support, sandboxing, evaluation, and observability packages.
Use Cases
- JavaScript teams building an agent that needs to call tools and run multi-step tasks, rather than just stream a single model response.
- Product developers adding chat to a React app, or bringing the same shared chat contract to another supported UI framework.
- Teams prototyping agents that need document retrieval, persistent memory, or provider switching.
- Developers creating terminal-based agents or exposing their tools through MCP-compatible clients.
What you need
Detected in the repository
- Node.js (from package.json)
- A test suite and automated checks on GitHub Actions
License in plain words
MITpermissive
- Commercial use: yes
- Modify and redistribute: yes
- You must keep: the copyright and license notice
- Share your changes: no
A summary, not legal advice: the LICENSE file is what applies.
Getting Started
Install the core and runtime, then follow the first-agent example in the README:
npm install @agentskit/core @agentskit/runtime tsx
Alternatives
- lagent: Lagent is a Python framework for composing agents and message-passing workflows, while AgentsKit is a TypeScript toolkit spanning adapters, tools, memory, retrieval, and UI.
- sandbase-harness: SandBase Harness centers on a local-first runtime with sandboxed tools, sessions, and audit and replay, while AgentsKit offers broader packages and shared contracts for building agents.
- agent-native: Agent-Native focuses on applications where people and agents share actions, data, and permissions, while AgentsKit provides a broader toolkit for agent runtimes and integrations.
| Project | Language | License | Stars | Status |
|---|---|---|---|---|
| AgentsKit | TypeScript | MIT | 36 | Active |
| lagent | Python | Apache-2.0 | 2.3k | Quiet |
| sandbase-harness | TypeScript | Apache-2.0 | 639 | New |
| agent-native | TypeScript | MIT | 7.5k | Active |
Considerations
The repository is active, with recent commits and releases, but its stability is package-specific: the README identifies @agentskit/core as stable and the other listed packages as beta. That breadth means adopters should check the stability policy and package-specific compatibility guidance before relying on a particular integration. The contributor base is small and highly concentrated, and the repository has a sizable open issue list, so teams should account for project continuity and support needs. Node.js is a stated runtime requirement; the README targets Node.js 20 or later.
Found this useful?
Share it with someone who would like AgentsKit.