Agent Group Chat
Bring agents with different roles into a group discussion; mention a member to invite a reply.
chat-groups
A native AI workspace
Give capable agents
a place to work.
Run capable agents with project context, tools, memory, skills, and reversible changes in one native desktop workspace.
Stable, RC, and Beta releases
Download for your computer.
WindowsWindows 10 / 11x64 full installer · WebView2 not bundled
macOSApple SiliconApple Silicon · .dmg
macOSIntel MacIntel processor · .dmg
LinuxDebian / Ubuntux64 package · .deb
LinuxFedora / RHELx64 package · .rpm
LinuxUniversal Linuxx64 portable app · AppImage
Select Stable, RC, or Beta to see its latest published build.
- Native
- A fast Tauri + Rust desktop app
- Inspectable
- See tool calls, approvals, diffs, and history
- Extensible
- Bring your own models, MCP tools, and skills
Build on OpenAgent
Your next capability starts here.
Follow the plugin tutorials from a small Skill to tools, a sidebar, and a verified release.
01
Your first plugin
Scaffold, validate, and install a Skill package.
02Package contract
Identity, compatibility, extensions, and persistent data.
03MCP tools
Add a server and choose Direct or Relay mounting.
04Sidebar & languages
Follow scoped context, host locale, and theme.
05Publish & update
Ship a stable archive and verify its update path.
Explore independently installed packages: Plugin Developer, Chat Groups, Goal, Graph, Cua Driver, and Message Board. Usage & plugin documentation →
Extend your workspace
Official plugin store.
Discover tools, workflows, integrations, and themes from the official OpenAgent catalog.
To install, open Integrations → Plugins in OpenAgent and find the plugin in the official store. Check its README for setup requirements.
Goal Execution
Keep an agent working toward a goal, track completed steps, and pause or resume progress.
goal
Task Orchestration
Split work into dependent tasks, run independent tasks in parallel, and view progress and results in a task graph.
graph
Desktop Control
Let an agent take screenshots, click, and type in desktop apps after you grant computer access.
cua-driver
Agent Message Board
Save agent collaboration messages in channels and threads, then search them or subscribe to updates.
message-board
Plugin Developer
Create OpenAgent plugins from requirements or templates, then develop, test, and validate them for installation.
openagent-plugin-kit
Asana Tasks
Find and update tasks and projects in Asana; requires an authorized Asana account.
asana
Context7 Docs
Look up library documentation and code examples with Context7 while writing code.
context7
Local Channel Test
Use a local web chat to test incoming messages, agent replies, and message-channel access rules.
fakechat
Firebase Projects
Inspect and manage Firebase project resources from your workspace; requires Firebase account access.
firebase
GitHub Repositories
Work with GitHub repositories, issues, and pull requests; requires a GitHub access token.
github
GitLab Projects
Work with GitLab repositories, issues, merge requests, and pipelines; requires GitLab authorization.
gitlab
iMessage Channel
Send messages to OpenAgent and receive agent replies through iMessage on a Mac; requires macOS message access.
imessage
Laravel Development
Inspect and debug the active Laravel app with Laravel Boost; requires PHP and Laravel Boost in the project.
laravel-boost
Langfuse Tracing
Send agent execution and tool-call traces to Langfuse after setup; message and model-content capture are separately opt-in.
langfuse
Linear Issues
Find and update issues and projects in Linear; requires an authorized Linear account.
linear
Playwright Browser Automation
Let an agent open web pages, click, fill forms, and take browser screenshots with Playwright.
playwright
Serena Code Navigation
Find symbols, follow references, and edit code by symbol with Serena in the active project.
serena
Terraform Infrastructure
Look up Terraform providers and modules and work with infrastructure workspaces; requires Docker, with account access for hosted workspaces.
terraform
Nova Compact Theme
A neutral gray theme with 13px text, rounded controls, and compact spacing; supports light and dark modes.
openagent-theme-nova
Vega Spacious Theme
A neutral gray theme with 14px text, rounded controls, and spacious spacing; supports light and dark modes.
openagent-theme-vega
Maia Rounded Theme
A neutral gray theme with 14px text, large rounded controls, and extra spacing; supports light and dark modes.
openagent-theme-maia
Lyra Monospace Theme
A neutral gray theme with 12px monospace text and square corners; supports light and dark modes.
openagent-theme-lyra
Mira Dense Theme
A neutral gray theme with 12px text, small rounded controls, and dense spacing; supports light and dark modes.
openagent-theme-mira
Luma Soft Theme
A neutral gray theme with 14px text, extra-rounded controls, and spacious spacing; supports light and dark modes.
openagent-theme-luma
Sera Warm Gray Theme
A warm taupe theme with 14px text and subtle rounded corners; supports light and dark modes.
openagent-theme-sera
Rhea Dense Rounded Theme
A neutral gray theme with 12px text, large rounded controls, and dense spacing; supports light and dark modes.
openagent-theme-rhea
Garden Emerald Theme
An olive-gray theme with emerald accents, violet charts, 14px text, and large rounded controls; supports light and dark modes.
openagent-theme-garden
Rose Ink Rose Theme
A warm taupe theme with rose accents, cyan charts, and 14px text; supports light and dark modes.
openagent-theme-rose-ink
No matching plugins. Try another search or category.
One workspace, the complete loop
More than a chat window.
OpenAgent combines an agent runtime, project tools, durable context, and desktop controls so useful work can continue across turns and sessions.
Agent runtime
Stream conversations, delegate sub-agents with Multi-Agent V2, and ask structured follow-up questions. Add Goal or Graph plugins for persistent workflows.
Tools & MCP
Apply patches, inspect images, run terminal commands, and connect external MCP servers over HTTP or stdio.
Memory & skills
Recall useful knowledge across sessions with hybrid search, and load reusable global or workspace-scoped SKILL.md instructions.
Desktop automation
Install the Cua Driver plugin for desktop automation, with host supervision and per-tool controls.
Messaging channels
Continue durable, scoped Agent conversations from Feishu/Lark, Telegram, QQ, WeChat, Discord, or Slack.
Checkpoints & branches
Search conversations, keep a durable checkpoint for every turn, branch from earlier messages, inspect diffs, and roll work back.
Rich, interactive output
Render streaming Markdown, code, Mermaid, ECharts, file and URL chips, plus image and video media.
Models & context
Connect Anthropic, OpenAI, or compatible providers, switch models, attach multimodal files, and compact long conversations.
Scheduled continuations
Ask the Agent to schedule a timed or completion-triggered continuation, then inspect or cancel pending hooks through its tools.
Workspace knowledge
Keep memory and skills scoped to a project or shared globally, with workspace DESIGN.md guidance.
Desktop experience
Reuse the existing primary window on repeated launches, keep workspaces in dedicated windows, and use native tray and file-manager controls.
Modular updates
Update the frontend, Runtime, and desktop shell independently, with separate plugin updates from verified GitHub releases.
Control without losing momentum
Autonomy,
with receipts.
Choose how independently agents can act while preserving the evidence needed to understand, approve, or undo their work.
Flexible approvals
Review every call, use model-assisted approval, disable approval, or apply the workspace sandbox.
Workspace boundaries
File and terminal actions stay aligned with the active project and its configured access policy.
Durable recovery
Conversation snapshots, branches, cancellations, and reverse diffs make experimentation recoverable.
Optional observability
Inspect local run details and optionally connect Langfuse tracing when deeper diagnostics are useful.
A clearer way to work
Context that
doesn't disappear.
- 01
Connect your model
Add an Anthropic, OpenAI, or compatible provider and choose the model that fits the task.
- 02
Open the right workspace
Select a project, attach relevant files, and let local memory, skills, and DESIGN.md supply focused context.
- 03
Watch and steer the work
Follow streamed output, tool calls, approvals, sub-agents, plugin workflow status, and file changes as they happen.
- 04
Keep or reverse the result
Review the diff, retain useful knowledge in memory or skills, branch to explore another path, or roll changes back.
Open source desktop AI
Make your next
agent task feel grounded.
Download the latest desktop release, connect a provider, and build a workspace that works the way you do.
Download the latest release