Introduction
Manage project work and run AI threads on a connected machine.
Convoy brings projects, tasks, and AI conversations into one web app. Use it to organize work with your team, keep decisions beside a task, and send work to an AI runtime on your own machine.
You can manage tasks without installing the CLI. To run AI threads, connect a machine with the Convoy CLI and an installed runtime. The CLI executes the work locally and streams the response into Convoy.
Pick your path
Use Convoy
Create a task, organize it in a List, and collaborate with your team.
Connect an AI runtime
Install the CLI, sign in through your browser, and send your first prompt.
Deploy for your team
Choose managed hosting or Docker, then configure authentication and your instance.
Contribute
Run the repository from source and follow the specs and testing workflow.
To work with Convoy tasks from an editor or agent, connect an MCP client. MCP works independently of the CLI bridge.
What Convoy includes
- Organizations and projects with explicit membership and permissions
- Tasks, Lists, comments, attachments, and Markdown or HTML artifacts
- AI threads with project agents, Skills, reusable prompts, and workflows
- A local CLI bridge for supported AI runtimes
- MCP tools and terminal commands for integrations and scripts
See supported runtimes and deployment status for runtime capabilities and deployment validation limits.
Learn the model
How Convoy works explains threads, profiles, and how a request reaches your machine. The FAQ answers common setup questions.