The manuals for these topics are maintained in the repositories listed below, so the text and the code move together. This site keeps the task-ordered entry points.

Component catalogue (48)

Component	Trust class	Purpose
`prose`	pure	Readable structured narrative
`callout`	pure	Important contextual warnings
`tabs`	interactive	Switch related presentations
`table`	interactive	Filterable structured reference
`code`	interactive	Syntax-highlighted copyable code
`code-diff`	interactive	Inspect a versioned example change
`live-code`	connected	Execute an example in an approved environment
`terminal`	connected	Interactive command example with sandbox binding
`api-console`	connected	Build and send approved API requests
`api-reference`	pure	Render canonical API schema and examples
`schema-explorer`	interactive	Explore an object or database schema
`graph`	interactive	Explore services and software relationships
`sequence`	interactive	Explain message ordering
`state-machine`	interactive	Explore a bounded state machine
`tutorial`	connected	Verified multistep task tutorial
`wizard`	connected	Collect configuration and validate prerequisites
`chat`	agentic	Page-scoped sourced documentation assistance
`chat-bubble`	agentic	Floating contextual documentation assistant
`troubleshooter`	agentic	Diagnose an authorized redacted error
`agent-task`	agentic	Run a scoped documentation task
`search`	interactive	Published content search with static fallback
`video`	interactive	Accessible recorded walkthrough
`gif`	pure	Short annotated motion with static alternative
`image`	pure	Accessible versioned screenshot
`release-timeline`	interactive	Published version and migration history
`migration`	interactive	Before-and-after change guide
`permissions`	interactive	Read-only permission matrix
`configurator`	connected	Prepare safe software configuration
`webhook-debugger`	connected	Inspect synthetic approved webhook delivery
`event-stream`	connected	Inspect an authorized event stream
`logs`	connected	View redacted approved diagnostic logs
`status`	connected	Read-only authorized deployment status
`feedback`	connected	Submit bounded document feedback
`checklist`	interactive	Track local tutorial progress
`cards`	pure	Choose task-oriented navigation
`breadcrumbs`	pure	Show page hierarchy
`version-selector`	interactive	Choose an actual available release
`locale-selector`	interactive	Choose an available language
`copy-context`	connected	Export permitted agent context
`apply-to-repo`	agentic	Prepare a reader-authorized integration PR
`embed`	connected	Isolated approved external UI
`metrics`	connected	Display authorized aggregate metrics
`media-transcript`	pure	Accessible transcript tied to media version
`faq`	pure	Task-oriented evidenced question answers
`glossary`	interactive	Software terminology and concepts
`download`	connected	Integrity-checked approved artifact download
`pricing-reference`	pure	Owner-sourced product prices; never inferred from code
`source-evidence`	interactive	Expose audience-safe verification evidence

Read the quickstart Compare approaches

Repository manuals

Core repository README — Prerequisites, install commands and the tested build/test flows.
API reference document — The written reference behind the generated operation list.
BYOK model credentials — Deployment BYOK and hosted per-tenant credentials, with the exact API paths.
Agent authoring — AGENTS.md/SKILL.md generation and the consent rules for agents.
OpenAPI and MCP bridge — How the OpenAPI documents become agent tools.
Component trust classes — What pure, interactive and connected components may do.
Component package format — The X1 manifest, digest and capability contract.
Market flow — Publish, discover, verify, install and render, with the evidence commands.
Self-hostable registry build — Building the static registry tree (index.json, llms.txt, sitemap).

Repository visibility is controlled by the project owner; where a repository is not public yet, the licence map and the published contracts remain the reference. Agents can read /llms.txt for a machine index of this host.