CLI Reference
Init & Clone
Every Spiral workspace begins by initializing a new project or cloning an existing one — just like Git.
spiral init
Run this inside any directory to give it a Spiral knowledge layer. The CLI scans your codebase, creates a project on the server, extracts context into structured facts, and writes IDE configuration files.
$ cd your-project
$ spiral init
Scanning repository...
Found Next.js, Tailwind, Prisma
Project created: your-project
Extracted 23 facts
$ spiral init
Scanning repository...
Found Next.js, Tailwind, Prisma
Project created: your-project
Extracted 23 facts
Files written to your project:
.spiral/config.json— Project identity anchor.cursor/mcp.json— Cursor MCP configuration.vscode/mcp.json— VS Code MCP configurationCONTEXT.md— Project knowledge summary.cursor/rules— Stack-specific AI rulesCLAUDE.md— Claude agent context.github/copilot-instructions.md— Copilot context.windsurfrules— Windsurf context
Use --manual to paste your own context instead of auto-scanning.
spiral clone
Download an existing knowledge base — either your own project from another machine, or a public knowledge base from the community.
$ spiral clone username/my-project
Fetching knowledge base...
Cloned into ./my-project (47 facts, 12 relationships)
Fetching knowledge base...
Cloned into ./my-project (47 facts, 12 relationships)
You can also clone by project ID: spiral clone <project-id>. The same IDE configuration files are generated automatically.