nevr context

Generate AI-optimized context from your Nevr application.

Usage

npx nevr context [options]

Description

This command generates a compact, AI-readable representation of your application's:

1. Entities - Names, fields, types, and constraints
2. Relations - How entities connect to each other
3. Plugins - Installed plugins and their endpoints
4. Extensions - Plugin fields added to other entities (e.g., username → user)
5. Actions - Custom entity actions
6. Access Rules - Who can do what
7. Semantic Fields - Searchable and embedding fields

The output is optimized for AI consumption, making it ideal for:

• AI coding assistants (Claude, Copilot)
• Documentation generation
• Codebase analysis
• AI-powered features

Options

Option	Alias	Default	Description
--config <path>	-c	./nevr.config.ts	Path to your config file
--output <path>	-o	context.md	Output file path
--format <fmt>		markdown	Output format: markdown or json
--json		false	Shorthand for --format json

Examples

Generate context.md (default):

npx nevr context

Custom output path:

npx nevr context -o ./docs/ai-context.md

Compact JSON format:

npx nevr context --json

JSON to file:

npx nevr context --format json -o ./context.json

Output Formats

Markdown (Default)

Human-readable format optimized for AI assistants:

# APP CONTEXT (Nevr)

Generated: 2024-01-15T10:30:00.000Z

## Entities

### user
- email (string) [required, unique]
- name (string) [required]
- role (string) [options:[admin,user]]

### post
- title (string) [required]
- body (text) [required, searchable]
- published (boolean)
→ author (belongsTo user, required)

Actions: publish, archive
Rules: list:[everyone] create:[authenticated] update:[owner]

## Plugins

### auth
Endpoints: signUpEmail, signInEmail, signOut, getSession
Entities: user, session, account

## Extensions (Plugin → Entity)

- **auth** → user: username, displayUsername
- **payment** → user: stripeCustomerId
- **organization** → session: activeOrganizationId, activeTeamId

## Searchable Fields
post.body, post.title

## Embeddings (RAG)
- post.body → openai/1536dim

JSON (Compact)

Minimal format for machine parsing:

{"v":"1.0.0","t":"2024-01-15T10:30:00.000Z","e":[{"n":"user","f":{"email":"s!U","name":"s!"},"r":[]},{"n":"post","f":{"title":"s!","body":"t!S","published":"b?"},"r":["author→user!"]}],"p":[{"id":"auth","e":["sign-up","sign-in"]}]}

Field encoding:

• s = string, t = text, i = int, b = boolean
• ! = required, ? = optional
• S = searchable, E = embedding, U = unique, X = sensitive

Use Cases

AI Coding Assistant

Add context to your project for AI assistants:

# Generate context file
npx nevr context -o ./CONTEXT.md

# Add to .cursorrules or similar
echo "See CONTEXT.md for application schema" >> .cursorrules

Documentation Generation

Use as input for doc generators:

npx nevr context -o ./docs/schema.md

RAG Integration

Generate context for vector search:

import { exec } from "child_process"

// Get context as string
const context = execSync("npx nevr context").toString()

// Store in vector database
await vectorStore.upsert({
  id: "app-context",
  content: context,
  metadata: { type: "schema" },
})

CI/CD Integration

Keep context updated automatically:

# .github/workflows/docs.yml
- name: Generate Context
  run: npx nevr context -o ./docs/CONTEXT.md

- name: Commit if changed
  run: |
    git add docs/CONTEXT.md
    git diff --staged --quiet || git commit -m "docs: update context"

Programmatic API

For more control, use the programmatic API:

import { generateContext, contextToMarkdown } from "nevr"
import { api } from "./src"

// Generate context object
const ctx = generateContext(api)

// Convert to formats
const markdown = contextToMarkdown(ctx)
const json = contextToJSON(ctx)

// Access data directly
console.log(ctx.entities)
console.log(ctx.plugins)
console.log(ctx.searchable)
console.log(ctx.embeddings)

See Also

• RAG - Retrieval-Augmented Generation
• nevr introspect - Show entities summary