Product Documentation

Detailed docs for how GTMship works.

These pages explain the core product surfaces in plain language: how workflows are authored, how connections are managed, how safe previewing works, and what happens when you deploy to your own cloud.

Start with Quickstart See the system flow
Workflow Builder Agent Custom Connections Agent Write Checkpoints AWS + GCP Deploy Persistent Memory Dashboard + CLI

What GTMship covers

GTMship is designed to take you from “I need a workflow that talks to these systems” to “that workflow is validated, connected, deployed, and observable in my own cloud account.” The platform combines a workflow SDK, a dashboard, a CLI, an auth service, a deployment engine, and memory so the product can keep state across sessions.

The docs here are focused on the product surfaces shown on the homepage. They explain what each feature does, what inputs it needs, and what GTMship manages for you under the hood.

Quickstart

01

Open the workspace

Launch GTMship locally and bring up the dashboard plus auth service so you can connect tools and iterate in Workflow Studio.

02

Connect your systems

Connect a built-in provider or give the Custom Connections Agent API docs for a custom service so it can configure auth and bindings.

03

Describe the workflow

Use the Workflow Builder Agent to explain the outcome you want. GTMship grounds API details before it generates or revises code.

04

Validate and preview

Run validation for runtime safety, then preview against sample input. Writes pause behind checkpoints when approval is needed.

05

Keep it local or deploy

Stay in a local-only loop for authoring, connection setup, validation, and preview. When the workflow is ready for always-on traffic, choose AWS Lambda or GCP Cloud Run and deploy it into your cloud account.

06

Monitor and iterate

Use logs, saved workflow state, and memory to keep improving the same automation instead of starting over each session.

gtmship dev
gtmship connections connect hubspot
gtmship validate
# deploy later when ready
gtmship deploy --provider aws

Local-only run

GTMship does not force cloud deployment on day one. A local-only run is enough for workflow authoring, connection setup, validation, and preview while you are still shaping the automation.

Good fit

Use local-only while the workflow is still taking shape

This is the fastest path when you are experimenting, testing credentials, previewing sample inputs, or proving the workflow with one operator at a time.

Deploy signal

Deploy when the workflow needs to stay on

Move to cloud deployment once you need stable webhook URLs, scheduled execution, shared production access, or a runtime that stays available beyond your local session.

Local-only does not mean a reduced product path. You still get Workflow Builder Agent, connections, validation, preview, and the shared dashboard plus CLI flow; you simply delay the cloud step.

What you bring and what GTMship manages

You bring

Outcome, systems, and operating rules

GTMship still needs the real business goal: what event should start the workflow, which systems it should talk to, what fields matter, and where human approval or business constraints should apply.

GTMship manages

Grounding, auth, safety, and deploy state

The product handles provider grounding, connection storage, validation rules, write checkpoints, artifact packaging, cloud deploy orchestration, and the status records needed to keep work visible after launch.

How GTMship works

Authoring

Ground first, then generate

The workflow agent is not treated like a one-shot code generator. It checks connected systems, pulls relevant docs or OpenAPI details, and only then drafts TypeScript plus a live flow representation.

Connections

Central auth service

Provider configs, encrypted credentials, OAuth callbacks, and connection bindings live behind the auth service so the dashboard and CLI can work from the same source of truth.

Safety

Validation and preview are separate

Validation checks workflow module shape and runtime rules. Preview simulates a run with sample input and pauses at write checkpoints when real-world mutations need approval.

Deployment

Deploy to your cloud

GTMship reads project config, resolves provider bindings, builds an artifact, deploys to AWS or GCP, and syncs deployment metadata back into its control plane.

  • SDK: defines workflow shape and runtime helpers.
  • Dashboard: interactive surface for connections, Workflow Studio, validation, preview, build, deploy, and logs.
  • CLI: project init, local runtime, provider management, connections, and deployment operations.
  • Deploy engine: packages and provisions cloud resources for supported targets.

Feature guides

Guide

Workflow Builder Agent

How Workflow Studio turns a prompt into grounded code, flow diagrams, validation, and next actions.
Guide

Custom Connections Agent

How built-in connectors and custom APIs are configured, authenticated, and bound into workflows.
Guide

Write Checkpoints

How preview mode pauses risky operations so you can inspect intended writes before live systems change.
Guide

Deployment

How GTMship plans bindings, packages artifacts, deploys to AWS or GCP, and syncs status back.
Guide

AI Memory

How saved context helps the agents reuse verified knowledge about your stack across sessions.
Guide

Dashboard + CLI

How the visual interface and terminal commands work against the same project, connections, and deployments.

Choose your next page

Start with workflow authoring Read about integrations See deployment flow
Docs live under /docs/ so each guide has its own shareable URL.