Clanker Docs

Appearance

Clanker CLIAI-Powered Cloud Infrastructure CLI

Open-source infrastructure agent for Clanker DevOps. For browser-based project work and reusable custom Agents, start with Clanker Cloud Workspace.

Get Started
Interactive Getting Started
Interactive Docs
Clanker Cloud
GitHub Repo

Natural Language Queries

Ask questions about your cloud infrastructure in plain English. Clanker translates your intent into the right API calls and CLI commands, then summarizes the results in a readable format.

Multi-Cloud Support

Work with AWS, GCP, Azure, Tencent Cloud, Cloudflare, DigitalOcean, Hetzner, Fly.io, Vercel, Railway, Verda, and Kubernetes from a single interface. Clanker automatically detects which provider you are asking about and routes your query accordingly.

Infrastructure as Code

Generate and apply infrastructure change plans using the maker pipeline. Clanker produces JSON execution plans that can be reviewed, versioned, and applied with a single command.

AI-Powered Agent System

An intelligent agent coordinator uses decision trees, dependency-aware parallel execution, and shared memory to handle complex multi-step investigations across your infrastructure.

Agent Sandboxes

Clanker Box describes and serves Cloud Run-ready sandboxes for the native Clanker CLI, Hermes, OpenClaw, Codex, Claude Code, and empty shell runtimes.

Collaboration Integrations

Query Sentry, Linear, and Notion directly or through natural language, with matching MCP tools for agent-driven incident, project, and knowledge-base workflows.

Kubernetes Management

Create, manage, and query EKS, GKE, and kubeadm clusters. Deploy workloads, inspect pod logs, gather resource metrics, and troubleshoot issues using natural language.

Cost Analysis

View real-time cost data across cloud providers. Get summaries, service breakdowns, daily trends, forecasts, anomaly detection, and tag-based cost allocation reports.

Looking for the browser-based Workspace? ​

Clanker Cloud Workspace keeps Project files, instructions, Sessions, and saved outputs together for people and AI agents. Read the AI agent workspace guide for persistent context and connected tools, or the AI agent factory guide to design, configure, test, maintain, and distribute custom Agents.

This documentation site focuses on the infrastructure CLI and desktop workflows. Hosted Workspace works in your browser; ordinary Project work does not require installing the CLI or configuring cloud-provider credentials. Compatible external assistants can use hosted Workspace MCP with browser authorization.