Our take
Activeregular commits and releases- Activity
- 90
- Community
- 72
- Issues
- 53
- Pull requests
- 84
+246 stars in the last 30 days
- Merges more pull requests than 95% of the projects we track
- Releases more often than 93% of the projects we track
- More contributors than 76% of the projects we track
Agent-Native is an actively developed framework for teams that want agents and a purpose-built interface to operate on the same application capabilities.
Good fit if
- You are building a TypeScript application where agent actions and React UI interactions should share implementation, validation, and permissions.
- Your product needs agent chat, persistent context, automations, or specialist agents alongside a user-facing interface.
- You can use the required recent Node.js runtime and want a project with regular commits and releases, contributions from 93 people, and no measured contributor concentration.
Look elsewhere if
- You need an agent that operates existing interfaces by clicking through them. Agent-Native instead has agents use the shared action layer.
- Your environment cannot run Node.js 22.22.0 or newer.
- You need a framework built around a language or UI stack other than TypeScript and React.
All health signals
| Last commit | 2026-10-25 (today) |
|---|---|
| Commits, last 90 days | 100+ |
| Releases, last 12 months | 100+ (latest @agent-native/toolkit@0.220.0, 2026-10-24) |
| Contributors | 93 (top contributor: 53% of commits) |
| Issues closed, last 90 days | 2+ |
| Pull requests merged, last 90 days | 93 (typically merged in 0 days) |
| Project age | 8 months |
Checked on 2026-10-25 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
7,736Forks
711Change since Nov 21, 2026: +6 stars·-1 forks
Overview
Agent-Native is a TypeScript framework for building agentic applications that pair autonomous work with a purpose-built UI. It addresses the gap between text-only agent interactions and knowledge work that benefits from visible context, inspection, editing, approval, and sharing. React interfaces and agents use the same action layer rather than having the agent operate the UI.
Actions share validation, permissions, and implementation across the UI and agent, and can also be exposed through HTTP, MCP, A2A, and the CLI. The framework includes agent chat, shared application data and state, authentication, skills and memory, automations, agent teams, and a PostgreSQL backend with PGlite support for local development.
Key Features
- Define capabilities as shared actions that agents call as tools and React code can invoke directly.
- Reuse action validation, permissions, and implementation across UI, agent, HTTP, MCP, A2A, and CLI surfaces.
- Share data and relevant application state between the agent and the UI.
- Build interfaces for delegating work and reviewing agent results.
- Add authentication, permissions, reusable skills, and persistent agent memory.
- Run agent work on schedules or events, or delegate work to specialist agents.
- Use PostgreSQL in production and PGlite for local development.
- Start from the included agent applications for tasks such as meeting notes, design, slides, analytics, calendar, mail, media, and content.
Use Cases
- A product team can build an agent that edits shared records while users inspect and approve changes in a React interface.
- Developers can expose an application capability to an agent and to HTTP, MCP, A2A, or CLI clients through one action definition.
- Teams can create specialist agents and scheduled workflows for recurring knowledge-work tasks.
- Developers can adapt the included apps as starting points for meeting, analytics, email, or content workflows.
What you need
Detected in the repository
- Node.js >=22.22.0 (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
npx --yes @agent-native/core@latest create my-agent --standalone --template chat
See the README and getting started guide for setup details.
Alternatives
- AgentsKit: AgentsKit provides a broader toolkit for agent runtimes, providers, memory, and retrieval, while agent-native centers on sharing actions between agents and React UIs.
| Project | Language | License | Stars | Status |
|---|---|---|---|---|
| agent-native | TypeScript | MIT | 7.5k | Active |
| agentskit | TypeScript | MIT | 38 | Active |
Considerations
This is a young project, but its measured health is active: it has regular recent commits and releases, contributions from 93 people, and no contributor concentration flag. The health data records only a small minimum of issues closed in the recent period, so it does not establish a typical issue-resolution pace. Production use involves choosing and operating a PostgreSQL backend; PGlite is offered for local development. The project is MIT licensed.
Found this useful?
Share it with someone who would like agent-native.