Clanker DevOps desktop app
Clanker Cloud is a browser-based AI workspace for people and agents. Its hosted Workspace keeps Project files, instructions, Sessions, and saved results together. It also supports an agent-factory workflow: configure custom Agents with selected knowledge, Skills, and Tools, test their saved work, and refine or share a reviewed version.
This page covers Clanker DevOps, the native desktop infrastructure workflow for macOS, Windows, and Linux. Use it to inspect, deploy, operate, and troubleshoot cloud infrastructure with AI agents and your locally configured provider access.
Choose where to start
| Your task | Starting point |
|---|---|
| Research, reports, documents, spreadsheets, Apps, or ongoing Project work | Open browser Workspace and read the AI agent workspace guide |
| Create, test, maintain, or distribute a reusable custom Agent | AI agent factory guide |
| Connect a compatible external assistant to hosted Workspace | Hosted Workspace MCP setup and permissions |
| Work with cloud infrastructure through the desktop app | Continue with the desktop capabilities below and desktop agent onboarding |
| Query infrastructure from a terminal | Clanker CLI documentation |
Ordinary hosted Workspace work does not require a desktop installation, cloud-provider credentials, or an infrastructure scan. A compatible external client connects to the hosted MCP endpoint through browser authorization; desktop MCP uses a separate local connection. Available tools depend on the client, account, organization settings, and permissions granted.
Desktop infrastructure capabilities
Deploy from GitHub to real environments
Point Clanker DevOps at a repo, review the inferred infrastructure plan, and approve the changes before applying them.
Talk to infrastructure in plain English
Ask what is running, what is failing, or where spend is climbing and get answers grounded in your connected environment.
Use local provider access
The desktop infrastructure workflow uses provider credentials configured on your machine and supports bring-your-own AI keys. This local credential model is separate from hosted Workspace processing. See the current data and privacy information for service processing and retention terms.
Work across providers from one surface
Work with AWS, GCP, Azure, Kubernetes, Cloudflare, Hetzner, DigitalOcean, and GitHub through the desktop infrastructure interface.
See topology, health, and cost together
Inspect infrastructure topology, runtime context, and cloud plus model spend alongside your investigation.
Connect local agents
The desktop workflow supports custom Node.js or Python agents running locally. Use the desktop app's local MCP connection when an external agent needs desktop infrastructure context. Follow desktop agent onboarding for that route. For hosted Projects and files, use hosted Workspace MCP.
Who uses the desktop workflow?
- Builders taking a repository into a real cloud environment.
- DevOps and SRE teams inspecting incidents and reviewing infrastructure changes.
- AI researchers working with their chosen cloud providers and local credentials.
For custom research, writing, reporting, or business Agents, start with browser Workspace and the agent factory guide.
Availability, pricing, and next steps
Use the current downloads page for desktop releases and supported platforms. Review pricing for current plans, credits, and usage rates, and the FAQ for current product capabilities. Account access, enabled tools, and connected environments determine which workflows are available.