Run all tests

Run all examples

Main test runner

Main test runner

Main test runner

Main entry point with argument parsing and mode selection.

Model information and capabilities example

Parse command line arguments.

Parse the documentation pages section to create sidebar navigation

Parse GitHub URL to extract owner and repository name. Handles various GitHub URL formats: - https://github.com/owner/repo - https://

Parse repository identifier into owner, repo, and branch components.

Parse the repository analysis section from README content

Show the difference between old and new API.

Print formatted subheader

Print a comprehensive test summary.

Process a chat message and generate AI response

Process repository from a Git URL, local directory, or a .zip file.

Process a chat message with streaming response - yields JSON strings

Read documents from directory

Generate new access token from refresh token

Reset daily count if it's a new day

Run comprehensive advanced features test suite

Run comprehensive user API key test suite

Run all tests

Run interactive test mode with user choices.

Convert repository identifier to filesystem-safe format.

Save the graph to a GraphML file using NetworkX.

Save the graph to a JSON file.

Save JSON data to file.

Save all repository files to storage. Args: user_id: User identifier repo_identifier: Repository identifier form

Save text content to file.

Save encrypted user API key with verification

Save wiki files to directory

Save ZIP content to file.

Combined search and context building

Search nodes by file path - optimized with index

Search repository context

Unzips a repository archive and flattens the directory structure by removing the top-level parent folder if one exists.

Simple chat method for quick usage Args: message: User message model: Model to use user:

Smart search that combines multiple search strategies

* Start streaming progress updates for a documentation generation task

Optimized streaming with better error handling

Stream chat response as JSON strings for FastAPI

Streaming response example

Structured output example

Test all advanced GraphSearchTool methods.

Test all available providers

Test API key availability from .env

Test API key encryption and decryption

Test different API key retrieval methods

Test basic streaming functionality

Test category and file-based searches

Test combination of streaming + structured output + function calling

Test context generation for LLM

Test multiple function calling

Test single function calling

Test loading graph data from JSON file

Test graph traversal operations (BFS, DFS, shortest path)

Test model capabilities and costs per user context

Test model configuration access

Test multi-term search functionality

Test multiple users with different API keys

Test performance of various operations

Test capabilities across different providers

Test automatic provider detection

Test all search operations

Test a single provider with its models

Test streaming functionality

Test streaming with function calling

Test complex structured output with nested schemas

Test simple structured output with JSON schema

Test the new subgraph-centric API.

Test user API key storage and retrieval (without database)

Test LLM calls using simulated user API keys