Skip to main content

Shipwright AI Documentation

Learn how to build, deploy, and operate software using Shipwright AI. Our documentation covers everything from initial setup to advanced infrastructure automation.

Note: Shipwright AI is in active development. This documentation describes both currently available features and planned capabilities. Items are clearly labeled where appropriate.

Getting Started

1Create an Account

Sign up at shipwright.in with your email or connect via GitHub/GitLab. Setup takes less than 60 seconds.

// Example flow:
Navigate to shipwright.in → Click "Join Waitlist" → Authenticate via OAuth

2Connect Your Repository

Link your GitHub repository. Shipwright AI will securely clone your code, analyze dependencies, and build a semantic graph of your application to provide context-aware AI assistance.

3Start Building

Once indexed, use the AI assistant to plan features, generate boilerplate, write tests, or explore architecture improvements. Everything stays synced with your codebase.

Concepts

Shipwright AI uses a specific hierarchy to organize your engineering workflows.

Workspaces

The highest-level container. Usually maps to your organization or company. Billing and global settings are applied here.

Projects

A specific application or service. Maps directly to a single Git repository. Contains your code context and deployment configurations.

Environments

Deployment targets for a Project (e.g., Development, Staging, Production). Environments securely store environment variables and infrastructure mappings.

Knowledge Graph

The underlying semantic index of your codebase that powers the AI. It maps dependencies, function signatures, and architectural patterns.

Projects

Projects are the core entity in Shipwright AI. Setting up a project correctly ensures the AI has complete context.

Project Configuration

A `shipwright.yml` file is automatically generated in the root of your repository to define build steps, AI behavior, and runtime specifications.

version: 1.0
project:
  name: my-webapp
  framework: nextjs
  language: typescript
ai:
  strict_mode: true
  test_framework: jest
deploy:
  build_command: "npm run build"
  output_dir: ".next"

AI Assistant

The AI Assistant acts as a senior engineer who knows your entire codebase. It operates in multiple modes:

  • Chat & Planning

    Use natural language to ask questions about your architecture, debug errors, or plan new features. The AI automatically pulls relevant context from your repo.

  • Inline Code Generation

    Generate components, refactor functions, or write unit tests. You can review diffs and apply changes directly to your branch.

  • Architecture Review

    Ask the AI to review a PR or branch against security best practices, performance guidelines, and your team's custom rules.

Deployment Planned

We are actively building deployment pipelines that integrate directly with your connected AWS accounts.

Planned Features:

  • Zero-downtime rolling deployments via AWS ECS / EKS.
  • Preview environments generated automatically for every Pull Request.
  • Instant rollback capabilities via immutable image tags.
  • Secure environment variable injection via AWS Secrets Manager.

Infrastructure Coming Soon

Shipwright AI will optionally manage your cloud resources using Infrastructure as Code (IaC) under the hood.

When you provision a database or caching layer in the UI, Shipwright AI will write and execute the corresponding Terraform/CloudFormation to your connected AWS account, ensuring you never lock your infrastructure into a proprietary platform.

API Reference Planned

A fully documented REST API will be available to programmatically manage projects, trigger deployments, and query AI models.

GET /v1/projects
POST /v1/projects/:id/deploy
GET /v1/deployments/:id/status
POST /v1/ai/completions

Authentication will be handled via Bearer tokens passed in the Authorization header.

CLI Planned

The Shipwright CLI brings the platform directly to your terminal. Interact with the AI assistant, manage environments, and view logs without leaving your editor.

$ npm install -g shipwright-cli

# Authenticate
$ shipwright login

# Ask the AI to explain a file
$ shipwright explain src/utils/auth.ts

# Trigger a deployment to staging
$ shipwright deploy --env staging

Authentication & Security

Shipwright AI uses industry-standard OAuth 2.0 / OIDC for all authentications. Single Sign-On (SAML/SSO) is planned for Business tier customers.

Data Protection

All data in transit is encrypted using TLS 1.3. Codebases and API keys stored at rest are encrypted with AES-256. See our Security Policy for details.

AI Privacy

We enforce strict data isolation boundaries. Your proprietary codebase is NEVER used to train foundational AI models.

Security

Security controls are applied at every layer of the platform. Shipwright AI is pre-launch, so we claim no certifications — what follows is what we actually implement today.

Encryption

Every connection is encrypted with TLS 1.3, and data stored in our databases and object storage is encrypted at rest with AES-256. Encryption keys are handled by a dedicated key management service.

Identity & Access

Authentication runs on OAuth 2.0 / OpenID Connect. Role-based access control (RBAC) governs what each user can do, and internal services follow the Principle of Least Privilege.

Secrets Handling

API keys, tokens, and credentials are never stored in source code. Secrets live in environment-based encrypted storage with controlled access, and our pipelines run secrets detection on every change.

Your Code and AI

Your proprietary code is never used to train AI models without explicit opt-in. You retain full ownership of your code and content, and third-party AI providers are bound by data processing agreements.

Found a vulnerability? Report it to security@shipwright.in and we'll acknowledge it within 48 hours. Our full practices are on the Security page, and the disclosure process is documented in our Security Policy.

Integrations

Shipwright AI is designed to integrate into your existing toolchain.

  • GitHub: Currently in prototype. Syncs repositories, reads PRs, and provides code context.
  • AWS: Currently available. Core infrastructure layer for deployments and storage.
  • Slack / Jira / Linear: Planned for Phase 2/3. For automated alerting, issue tracking, and team visibility.

Visit our Integrations page for the full list.

Troubleshooting

Repository Connection Fails

Ensure you have granted the Shipwright AI OAuth app access to the specific repositories you want to connect. Organization repositories require explicit admin approval.

AI Missing Context

If the AI suggests code that ignores your existing project patterns, force a manual re-index of the knowledge graph in your Project Settings.

Frequently Asked Questions

What is Shipwright AI?
Shipwright AI is an AI-powered developer and DevOps platform that helps software teams plan, build, deploy, monitor, and maintain applications from a single unified interface.
Who is Shipwright AI for?
Early-stage startups, small software companies, independent development teams, DevOps teams, and full-stack developers building cloud applications.
How does the AI assistant work?
The AI assistant connects to your code repositories, understands your architecture and dependencies, and provides contextual assistance for code generation, testing, documentation, and planning.
Does Shipwright AI replace developers?
No. Shipwright AI amplifies developer capabilities by automating repetitive tasks, providing contextual suggestions, and reducing context switching. Developers make all critical decisions.
Which cloud providers are supported?
We are initially focusing on AWS integration. Support for Google Cloud and Microsoft Azure is under evaluation.
Does Shipwright AI work with GitHub?
Yes. GitHub integration is a core feature currently in prototype. GitLab support is planned.
Is there a REST API?
A REST API is planned and will be available at launch. The API will be documented with an OpenAPI specification.
Is there a free plan?
We intend to offer a free Developer tier for individuals and open-source projects. Final pricing has not been determined.
How is my code protected?
All data is encrypted in transit (TLS 1.3) and at rest (AES-256). We follow the Principle of Least Privilege and do not use your proprietary code to train AI models without explicit opt-in.
Where is data stored?
Data is stored on AWS infrastructure. Specific region information will be published at launch.
Can teams collaborate?
Team collaboration features including shared projects, engineering notes, task management, and activity history are planned for Phase 5 of our roadmap.
What integrations are available?
GitHub, AWS, and Docker integrations are available or in prototype. Slack, Jira, Linear, GitLab, and others are planned.
What is on the roadmap?
Our roadmap covers five phases: Foundation, Developer Tooling, Cloud Automation, Operations, and Scale. See our Roadmap page for details.
How do I join early access?
Visit our Contact page and submit the waitlist form. We'll reach out as early access spots open up.
How can I contact the team?
Email hello@shipwright.in or use the contact form on our Contact page.

Need help?

Can't find what you're looking for? Our engineering team is here to help you get the most out of Shipwright AI.

Contact Support