Context Scanner

AI-Friendly Codebase Documentation Generator

AI coding assistants are powerful but context-blind. They don't know your architecture, your conventions, or your compliance requirements until you explain them—every single time. Context Scanner solves this by generating structured documentation that AI assistants can parse instantly.

The Context Problem

Every time you start a new AI coding session, you're back to square one. The assistant doesn't remember your project structure. It doesn't know you use Tailwind with a custom design system. It doesn't know you have NDIS compliance requirements.

You waste the first 10 messages establishing context. Then you do it again tomorrow. And the day after.

The question: What if the AI could understand your entire codebase the moment it opened the project?

9 Context Files

One command generates a complete context package. Each file is optimised for AI parsing while remaining human-readable.

1context-summary.md

Overview for AI assistants - start here

2architecture.md

Project structure, tech stack, design patterns

3api-reference.md

API endpoints, functions, components

4data-models.md

Prisma schema, TypeScript interfaces

5validation-rules.md

Zod schemas, environment variables

6conventions.md

Coding standards, naming patterns

7integration-map.md

External services, dependencies

8compliance.md

Security, accessibility, privacy gap analysis

9status.md

Project health, TODOs, recent activity

Compliance Gap Detection

The scanner automatically detects compliance indicators across Australian frameworks. It doesn't just document what exists—it identifies what's missing.

Security

OWASP Top 10 · ACSC Essential Eight

Rate limitingCSRF protectionAuth patternsAudit loggingEncryptionInput sanitisation

Accessibility

WCAG 2.1 AAA · HX Framework

ARIA attributesSemantic HTMLAlt textFocus managementScreen reader support

Privacy

OAIC APP 1-13 · Australian Privacy Principles

PII handlingConsent trackingData retention policiesCross-border data handling

AI Governance

VAISS 10 Guardrails · PSB Protocol

Bias detectionHuman review workflowsExplainability featuresContestability mechanisms

Federated Mode

For microservices or multi-repo architectures, federated mode solves context fragmentation. It generates cross-service documentation that shows how services connect.

The Problem

xCode distributed across repositories
xTypes and contracts may drift
xCross-service impacts invisible
xAI can't see the whole picture

Federated Solution

+Shared Types: Types that appear in multiple services - potential contracts
+API Integration Points: Functions that make HTTP calls to other services
+Database Schema Overlap: Same model names across services - potential coupling
+Event Contracts: Event-driven patterns (emit, publish, subscribe)
+Shared Dependencies: NPM packages used by multiple services

Interactive Architecture

Explore how Context Scanner processes your codebase. Click on any node for details, and use the flow buttons to follow data through the pipeline.

Usage

Simple CLI interface. Run once, get instant context.

Single Service

# Scan current directory
context-scanner

# Scan specific directory
context-scanner ./my-project

# Custom output directory
context-scanner -o ./docs/context

Federated Mode

# Scan multiple services
context-scanner -f ./api ./web ./worker

# Custom federation output
context-scanner -f ./api ./web -o ./docs

Real Results

Example output from scanning the DSX Platform codebase:

801

Source files

984

Functions

305

Types/interfaces

Database models

Validation schemas

158

Environment vars

Technical Stack

TypeScript

Node.js

AST Parsing

Glob Patterns

Prisma Schema Parser

Zod Detection

WCAG Analysis

Markdown Output

Want Context Scanner for Your Codebase?

Context Scanner is currently used internally at aXai. If you're interested in using it for your projects or want to discuss how it could help your team, get in touch.

Get in Touch View All Innovations