stagewise
Run coding agents with cloud, external, or local models
Juggler
Inspect, branch, and manage coding-agent conversations across providers.
Meet Juggler
Juggler is an open-source interface for coding agents that makes their conversations and model requests inspectable. Users can examine prompts, tool arguments, results, token usage, and timing, then branch, rewind, or edit a session. Recursive subthreads keep experiments and delegated work separate from the main conversation.
The app connects directly to supported providers, existing agent subscriptions, or local models. Multiple agents can work in separate worktrees or other isolated workspaces, while command approval rules remain editable. Juggler can run locally as a desktop app or beside code on a server, with clients sharing the same session. Conversations and API keys stay on the host machine.
Discussion
Tried it? Share a useful thought or ask a question.
Good conversations start with a hello.
Sign in to commentProduct details
- Website
- juggler.studio
- Status
- Launched
- Category
- Developer tools
- Listed
Explore related topics
Similar products
Lucid Train
Turn codebase architecture into a visual specification for coding agents.
Sente
A terminal coding agent connected to the teai.io model catalog.
OpenPilot
An open-source desktop AI agent that uses your own models.
Invoke
Plan, build, debug, and review code with configurable agents
OpenCode Superapp
Run coding models, voice, and desktop tools inside an extensible OpenCode workspace.