Core Features

At its heart, API Forge is designed to make your life easier. It breaks down the complex task of API integration into two simple, automated phases that feel intuitive and powerful.

Phase 1: Let the Discovery Begin

Your journey starts here. Instead of manually hunting for endpoints, you just need to point API Forge in the right direction. It handles the rest.

Automated Crawling

Provide a single entrypoint URL (like a 'Getting Started' page), and API Forge will fetch the page, find all relevant links, and fetch their titles.

Intelligent Ranking

The tool uses AI to analyze the list of discovered pages, categorizing them by method (GET, POST, etc.) and recommending the most fundamental endpoints to start with.

User Selection

You are presented with the categorized and ranked list, allowing you to select which endpoints to process.

Phase 2: From Analysis to Action

With your endpoints selected, this is where the real transformation happens. Our AI agents get to work, turning documentation into tangible assets for your project.

Deep Analysis

For each selected endpoint, a specialized AI agent parses its documentation to extract critical details: the HTTP method, request parameters, and response schema.

Cumulative Schema

The application uses the analysis to build or update a single, comprehensive database schema, which is rendered as a fully interactive diagram on the Schema Canvas. You can rearrange tables, inspect attributes on hover, and customize the appearance.

Mock Function Stubs

Ready-to-use JavaScript functions with rich JSDoc comments are generated for each endpoint, providing a clear starting point for your implementation.

Advanced Schema Management

The schema is your playground. The Schema Canvas is more than just a viewer—it’s a fully interactive environment designed to give you complete clarity and control over your data model. We believe in empowering you with tools that are both powerful and intuitive.

Interactive Canvas Features

The canvas is more than just a static image. You have full control over the presentation of your schema to make it as readable as possible.

Click & Drag

Rearrange tables by clicking and dragging them to your desired position. The layout will adjust automatically to the new positions.

Hover to Inspect

To keep the diagram clean, nodes only display table titles. Hover over any table for one second to reveal a tooltip with its full list of attributes.

Intelligent Highlighting

Click on any table or relationship to highlight it and its immediate neighbors, fading out the rest of the diagram for focused analysis.

Customizable Accent Color

Use the palette icon in the top-right corner of the canvas to change the highlight color to your preference from a selection of themes.

Smart Entity Shapes

The visualizer automatically detects junction tables (tables with two or more primary/foreign keys) and renders them as diamonds to distinguish them from standard entity tables.

Reset View

If you get lost while panning or zooming, use the reset button in the top-right corner to instantly center the diagram and restore the default zoom level.

Interaction History & Debugging

We believe in transparency. The Interaction History page, accessible via "Admin Mode," gives you a behind-the-scenes look at every step the system takes. It’s your command center for understanding, debugging, and gaining confidence in the generation process.

Understanding the History Table

The main view is a table that logs every high-level action (e.g., "Discover") and its underlying sub-actions (e.g., "Fetch Documentation", "Rank Endpoints"). Each row provides key details:

Action & Status

The name of the operation being performed and whether it was a Success, an Error, or is still In Progress.

Input & Details

A summary of the data the action started with and what it accomplished. You can hover for more details.

Duration & Timestamp

How long each action took, helping you identify performance bottlenecks. Hover over the timestamp for precise start and end times.

Debugging with Flow Details

The most powerful feature for debugging is the ability to inspect the exact inputs and outputs of the underlying AI flows. Click the chevron icon at the end of any row to expand it and see the raw JSON data.

Troubleshoot Errors

If an AI flow fails, you can see the exact input that caused the issue, helping you diagnose malformed or missing documentation content.

Improve Prompts

By examining the AI's output for a given input, you can better understand its reasoning and identify ways to improve the system prompts in the source code.

// Example of Flow Input/Output for a failed schema fix
{
  "flowInput": {
    "brokenSchema": "erDiagram User }|..|{ Post : writes"
  },
  "flowOutput": {
    "fixedSchema": "erDiagram
  User ||--o{ Post : "writes""
  }
}