# Tapstate documentation

> tapstate is a unified operational data engine that builds and maintains an Operational State Layer.
>
> Its product model captures changes, transforms records in flight, and delivers current operational state to applications, APIs, automation, and AI agents through one governed Capture–Transform–Serve data path. It is intended for teams that would otherwise assemble separate CDC, broker, stream-processing, and serving products for the same path. This describes the product model; it does not mean every surface is available in the current preview.
>
> The current preview is a prerelease, single-node, in-memory runtime. The documented end-to-end path captures MySQL orders and PostgreSQL shipments, assembles them with `nest`, and maintains the result in MongoDB. Use the linked page-level documentation for current maturity, roles, modes, limitations, and setup requirements.

## When to use tapstate

- Keep operational state current for applications, APIs, automation, or AI agents.
- Reduce the coordination cost of operating separate CDC, broker, stream-processing, and serving systems.
- Maintain a reviewable Capture–Transform–Serve path across source and target systems.
- Offload reads, migrate while a source remains active, or publish governed operational feeds.

See [Use cases](https://tapstate.dev/docs/overview/use-cases) and [tapstate vs. a streaming stack](https://tapstate.dev/docs/overview/vs-streaming-stack) for tradeoffs and verification questions.

## Start here

- [What is tapstate?](https://tapstate.dev/docs/overview/what-is-tapstate)
- [Quickstart](https://tapstate.dev/docs/overview/quickstart)
- [Explore the order-state demo](https://tapstate.dev/docs/overview/order-state-demo)
- [Install the CLI](https://tapstate.dev/docs/overview/install)
- [Create your own workspace](https://tapstate.dev/docs/overview/create-workspace)
- [Architecture](https://tapstate.dev/docs/overview/architecture)

## Connector status

- Current product path: [MySQL source](https://tapstate.dev/docs/connectors/mysql) and [PostgreSQL source](https://tapstate.dev/docs/connectors/postgresql) with Snapshot and CDC into a [MongoDB target](https://tapstate.dev/docs/connectors/mongodb). The Quickstart assembles related rows from both sources with `nest`.
- Other connector preparation materials are not published in the public documentation site or combined agent context and do not imply a product contract.

## Agent guidance

- Prefer page-level Markdown for implementation decisions; this file is a discovery index.
- Treat connector frontmatter as compact metadata and use the page body for the canonical explanation.
- Check maturity, roles, modes, limitations, and setup guidance together.
- Do not infer unavailable UI or runtime behavior from upstream connector documentation.

## Documentation map

- [What is tapstate?](https://tapstate.dev/docs/overview/what-is-tapstate)
- [Quickstart](https://tapstate.dev/docs/overview/quickstart)
- [Explore the order-state demo](https://tapstate.dev/docs/overview/order-state-demo)
- [Install the CLI](https://tapstate.dev/docs/overview/install)
- [Create your own workspace](https://tapstate.dev/docs/overview/create-workspace)
- Understand tapstate
  - [Use cases](https://tapstate.dev/docs/overview/use-cases)
  - [Architecture](https://tapstate.dev/docs/overview/architecture)
  - [Connections and connectors](https://tapstate.dev/docs/concepts/connections-and-connectors): Understand how connector definitions become reusable source and target connections
  - [Capture modes](https://tapstate.dev/docs/concepts/capture-modes): Choose how a tapstate source reads existing data and later changes
  - [Resources and pipelines](https://tapstate.dev/docs/concepts/dsl): Understand the tapstate resource model and how reusable declarations compose into a pipeline
  - [State and storage](https://tapstate.dev/docs/concepts/storage): Understand the state a tapstate data path must retain for current reads, recovery, and audit
  - [tapstate vs. a streaming stack](https://tapstate.dev/docs/overview/vs-streaming-stack)
- Connectors
  - [Connector guides](https://tapstate.dev/docs/connectors): Configure the current MySQL and PostgreSQL source paths into MongoDB operational state
  - [MySQL](https://tapstate.dev/docs/connectors/mysql): Capture an initial MySQL load and ongoing changes into tapstate, then maintain application-ready operational state in MongoDB.
  - [PostgreSQL](https://tapstate.dev/docs/connectors/postgresql): Capture an initial PostgreSQL load and ongoing changes into tapstate, then maintain application-ready operational state in MongoDB.
  - [MongoDB](https://tapstate.dev/docs/connectors/mongodb): Materialize MySQL snapshots and ongoing changes as application-ready operational state in a MongoDB target.
- Build and operate

  - **Set up and run**
  - [Bootstrap and authenticate](https://tapstate.dev/docs/guides/bootstrap-and-auth): Create the first tapstate administrator, sign in, and use the preview control plane safely
  - [Register and test connectors](https://tapstate.dev/docs/guides/register-connectors): Register connector JARs, apply connection resources, test connectivity, and discover source schemas
  - [Apply and run a pipeline](https://tapstate.dev/docs/guides/apply-and-run): Apply a validated workspace and control the tapstate preview pipeline lifecycle

  - **Document assembly**
  - [Assemble documents with nest](https://tapstate.dev/docs/guides/assemble-documents-with-nest): Build and maintain one document from related MySQL and PostgreSQL tables
  - [Handle structural key changes with nest](https://tapstate.dev/docs/guides/nest-structural-key-changes): Keep assembled documents correct when a root, parent, or array key can change
  - [Plan nest capacity and delivery behavior](https://tapstate.dev/docs/guides/nest-throughput): Size a nest pipeline around whole-document writes, final-state delivery, and representative source changes

  - **Operate and troubleshoot**
  - [Observe a pipeline](https://tapstate.dev/docs/guides/observe-a-pipeline): Interpret tapstate pipeline status, metrics, snapshot progress, and node-local logs
  - [Troubleshoot a tapstate configuration](https://tapstate.dev/docs/guides/troubleshooting): Isolate authoring, connector, deployment, and data-correctness problems without confusing one layer for another
- Reference

  - **Interfaces**
  - [CLI reference](https://tapstate.dev/docs/reference/cli): Use tapstate offline workspace commands, connected sessions, and one-shot commands
  - [MCP integration](https://tapstate.dev/docs/reference/mcp): Connect an MCP-compatible host to a running tapstate server through the CLI gateway
  - [REST API preview](https://tapstate.dev/docs/reference/rest-api): Call the authenticated tapstate control-plane endpoints and interpret preview responses

  - **Workspace format**
  - [Resource grammar](https://tapstate.dev/docs/reference/dsl-grammar): Reference for tapstate resource files, relationships, and validation boundaries
  - [DSL fields reference](https://tapstate.dev/docs/reference/dsl-fields): Schema-generated field lookup for the tapstate/v1 resource contract

  - **Resource definitions**
  - [source](https://tapstate.dev/docs/reference/source): Define a reusable tapstate source connection, including its connector ID, endpoint settings, and source read mode.
  - [pipeline](https://tapstate.dev/docs/reference/pipeline): Define a tapstate pipeline that selects source data, applies transforms, and materializes or delivers the result.
  - [transforms](https://tapstate.dev/docs/reference/transforms): Reference for inline and reusable transform shapes
  - [view and serve](https://tapstate.dev/docs/reference/view-and-serve): Reference for reusable view materialization and serve surface resources
- Release notes
  - [v0.4](https://tapstate.dev/docs/releases/v0.4): Install the v0.4 preview CLI, resume saved sessions, diagnose sync startup, and recover interrupted downloads.
  - [v0.3](https://tapstate.dev/docs/releases/v0.3): Explore cross-source operational state, PostgreSQL capture, managed views, and data inspection in the v0.3 preview line.
  - [v0.2](https://tapstate.dev/docs/releases/v0.2): Review the tapstate v0.2 preview capabilities, nest transforms, playground setup, and known runtime limitations.
  - [v0.1](https://tapstate.dev/docs/releases/v0.1): Understand what you can explore in the tapstate v0.1 preview and its local MySQL-to-MongoDB demo.
- AI & agents
  - [AI-readable documentation](https://tapstate.dev/docs/for-ai/llms): Canonical tapstate documentation surfaces for search engines, agents, and language models
  - [Use tapstate docs with AI](https://tapstate.dev/docs/for-ai/authoring): Give an AI assistant canonical connector and resource context, then review and validate its output