Store and manage reusable documentation that helps AI generate code matching your standards and requirements.

Knowledge Base

The Knowledge Base stores reusable documentation that helps AI generate code matching your standards and requirements. This guide covers creating, managing, and using Knowledge articles.

What Is the Knowledge Base?

The Knowledge Base contains text articles with information AI uses when generating code:

Coding standards and style guides
API documentation for internal or external services
Architecture documentation explaining system design
Technical specifications for features or integrations
Framework guidelines for libraries you use

When you attach Knowledge articles to a Work Item, AI reads them and follows the guidelines when generating code.

Accessing the Knowledge Base

Click Knowledge in the sidebar to open the Knowledge Base editor.

The editor has two panels:

Left panel: List of Knowledge articles organized by scope
Right panel: Markdown editor for article content

Creating Knowledge Articles

Create a New Article

Click Create next to the Organization or Project section header
Enter a name for your article (e.g., "React Component Guidelines")
Write your content in the markdown editor
Click Save in the toolbar

Import from Markdown Files

Drag a .md file into the editor
AngenAI creates a new article with the filename as its name
Edit the name and content as needed
Click Save

Article Scopes

Knowledge articles have two scopes:

Organization-Wide

Articles available across all projects in your organization.

Use for:

Company-wide coding standards
Shared API documentation
Common architectural patterns
Universal style guides

Note: Only Admins and Owners can create or modify organization-wide articles.

Project-Scoped

Articles available only within a specific project.

Use for:

Project-specific conventions
Local API endpoints
Feature-specific documentation
Project architecture notes

Members can create and modify project-scoped articles.

Writing Effective Knowledge Articles

Be Specific and Actionable

Provide clear guidelines AI can follow:

Good example:

## React Component Structure

All React components must:
1. Use functional components with hooks
2. Define TypeScript interfaces for props
3. Export components as named exports
4. Place styles in a separate `.module.css` file

Example:
\`\`\`tsx
interface ButtonProps {
  label: string;
  onClick: () => void;
}

export function Button({ label, onClick }: ButtonProps) {
  return <button onClick={onClick}>{label}</button>;
}
\`\`\`

Include Examples

Show code examples that demonstrate correct implementation:

## Error Handling

Use try-catch blocks with typed error responses:

\`\`\`typescript
try {
  const result = await userService.create(data);
  return { success: true, data: result };
} catch (error) {
  logger.error('User creation failed', { error, data });
  throw new ServiceError('USER_CREATE_FAILED', error);
}
\`\`\`

Document Patterns

Explain patterns and when to use them:

## Repository Pattern

Use the repository pattern for database access:

- Create a repository interface defining methods
- Implement the interface with database-specific code
- Inject repositories into services via dependency injection

This allows swapping databases without changing business logic.