Himalayan village and mountain landscape

Jade developer docs

Build with the same Jade product surface.

Public API references, integration guides, and agent-ready instructions using the same navigation and footer as the homepage.

Documentation

Build content-backed products with Jade.

A practical path from content model to production delivery, built for developers, editors, browser clients, server integrations, and coding agents.

Start with the APIGive this to an agent

Start

The happy path is short: model content, publish it, mint a key, then fetch published records from the public API.

1

Model content

Create a project, collection, and active fields in the Jade app.

2

Publish entries

Draft as much as needed. The public API returns published documents only.

3

Create an API key

Add allowed browser origins, or leave origins blank for server-only prototypes.

4

Fetch content

Use stable project and collection slugs in URLs.

First request

curl \
  -H "Authorization: Bearer $JADE_API_KEY" \
  https://jade.kabeli.org/api/public/v1/projects

Mental model

Jade keeps editing workflows private and delivery workflows narrow. That separation is what makes the integration predictable.

Workspace scoped

API keys belong to one workspace. Scopes decide whether a key can read delivery content, submit forms, or manage schemas, documents, and assets.

Separate delivery and writes

Public delivery uses read and form scopes. Trusted integrations use management scopes for collection, field, document, and asset CRUD.

Origin-aware browser delivery

Allowed origins decide which browser domains can fetch with a key. Server requests are not blocked by browser CORS.

Integration paths

Pick the path that matches where the content is rendered.

API referenceEndpoint reference, authentication, CORS, schemas, errors, and production examples.AI agentsPrompt-ready context, code-generation rules, and guardrails for agentic coding workflows.Storage and assetsConnect Cloudflare R2, configure delivery URLs, and use the project asset library.

AI and agentic coding

The docs are written for humans, but also expose a concise machine-readable contract for tools that consume documentation.

Use the agent contract

Send coding agents to /docs/agents, /llms.txt, or /llms-full.txt. The contract tells them which endpoints are public, how keys are passed, what not to call, and how to keep secrets out of client bundles.

Production checklist

Store API keys in server-side environment variables when possible.
If a browser must call the API, configure exact allowed origins.
Use published documents for production rendering; keep draft previews private.
Paginate collection reads with limit and offset.
Roll keys after accidental exposure. Lock keys when only superadmins should change them.