Quick Start
Welcome to Anydocs
Quickly understand what Anydocs is and which path to start with on your first visit.
Anydocs puts Studio, CLI, reader builds, and AI collaboration into one docs-project model. On your first visit, choosing the right path matters more than memorizing every term.
Who this site is for
Treat this documentation set as the entry point for three main paths.
- Content authors who want to create, edit, organize, and publish pages
- Site maintainers who need reader builds, search indexes, and AI artifacts
- AI collaborators who want Claude Code, Codex, or similar agents to work through MCP
The three delivery surfaces in Anydocs
Whether the work is human-authored or AI-assisted, everything lands in the same project model.
- Studio: visual editing for pages, metadata, navigation, and project settings
- CLI: project initialization, reader preview, static builds, and legacy imports
- AI delivery: MCP for canonical reads and writes, plus `llms.txt`, `llms-full.txt`, and `mcp/*.json` for machine-readable outputs
Recommended reading paths
Choose the shortest path for your goal instead of reading the whole site linearly.
- Manual authoring path: Installation -> First Documentation Project -> Open Project -> Edit Page -> Manage Navigation
- AI collaboration path: Installation -> AI Authoring Workflow -> Claude Code / Codex integration -> MCP Tools Reference
- Legacy migration path: CLI import -> convert-import -> Publishing Workflow -> Build Output
The rule to remember before anything else
Public visibility in Anydocs is controlled by page status.
- Only `published` pages enter the reader, search indexes, `llms.txt`, `llms-full.txt`, and `mcp/` outputs
- `draft` and `in_review` pages remain in the source project but do not enter public outputs
- An AI authoring task is not complete when the draft is written; it is complete when the content crosses the right publication boundary
Pick a path first: If you only need to get a docs site running, start with Installation and First Documentation Project. If you want an agent to participate in authoring, continue with AI Authoring Workflow.