Introducing AI tools for building on Veracity

Veracity now provides a dedicated AI hub: a home for the tools and guidance that let AI assistants do more of the context-gathering, grounded in official, verified Veracity documentation.

Context

Building on Veracity means making several decisions correctly in the right order: registering an application, wiring authentication through the Veracity Identity Provider, connecting to platform APIs, and meeting the requirements for production approval. Each step has its own documentation, its own configuration, and its own failure modes.

For most builders, that means constant movement between the editor and a browser: searching for the right guide, checking a parameter, copying an example, then returning to the code. The work itself is not difficult, but the overhead accumulates. And unlike generic API integration, Veracity integrations run in industrial and maritime contexts where incorrect access control or authentication has operational consequences.

What is available now

The Veracity Development MCP Server is the centrepiece. An MCP server is a service that exposes tools, prompts, and resources to an agent over the Model Context Protocol. Add the Veracity Development MCP Server to your AI coding assistant and it can search and retrieve official Veracity documentation at the time you ask: not from training data alone, but from the live documentation. The code your assistant produces reflects current platform behaviour.

Setup guides are available for GitHub Copilot, Cursor, Claude Code, Codex, OpenCode, and Windsurf. Each guide covers configuration, project instructions, and verification in under five minutes.

Agent skills are instruction files you place in your repository. A skill is a Markdown file your agent loads when it recognises a relevant task and applies to its output. Skills for common Veracity integration tasks are currently in internal preview and will be released publicly once stable.

Prompts and project rules are copy-pasteable instruction files that establish Veracity conventions for every session in your repository (correct OAuth flows, explicit tenant context, credential handling rules) without requiring you to repeat them in every prompt.

Working safely covers the areas where AI-assisted Veracity development requires deliberate care: credentials, regulated maritime data, multi-tenancy, code verification, and testing before production.

Use cases

Starting a new Veracity integration: add the MCP server to your repository. Ask your agent to implement the OAuth flow for your application. The agent retrieves the current Veracity Identity Provider documentation and generates code that follows the documented patterns. You review it against the verification checklist and test it against development credentials.

Reviewing existing integration code: paste the code review prompt from the prompts page into your agent session. The agent checks endpoint paths, parameter names, tenant handling, and credential patterns against current Veracity documentation and flags discrepancies.

How to get started

1. Add the Veracity Development MCP Server to your AI coding assistant. This takes under five minutes.
2. Add the baseline prompt to your repository so Veracity conventions apply to every session.
3. Check Agent skills for preview skills that may already cover your use case.

Everything you need to get started is in Build with AI on Veracity: setup guides for each supported agent, prompt templates, and working safely guidance.