List all OAuth2 clients.

List all API keys.

List all nodes and their capabilities.

Get list of configured server names

Load character presets from YAML file.

Load all text files from the specified directory. Args: directory_path: Path to the directory containing text files file

Log in to Gmail using the specified browser and credentials. :param browser: Browser to use (e.g., "Safari", "Chrome", "Fire

Main function to handle stdin/stdout interface.

Main application entry point

Main function to demonstrate the document search with different retriever types.

Main Streamlit application.

Main entry point.

Main entry point.

Main test execution.

Main test execution.

Main entry point.

Main entry point.

Main test runner

Main example function.

Main example function.

Click a menu item in an application menu bar. :param app_name: Name of the application :param menu_name: Name of the

Example of multi-turn conversation.

Handle multimodal chat completions using the Qwen2.5-Omni model. Args: messages: List of message dictionaries with 'role

Handle multimodal chat completions using the Qwen2.5-Omni model. Args: messages: List of message dictionaries with 'role

Open a macOS application using the 'open' command.

Find the optimal num_tasks_per_round value by running experiments Args: task: Task for minions to perform

Extract clean markdown content from URLs using Parallel AI's Extract API. Converts any public URL into clean markdown, including JavaScr

Search the web using Parallel AI's Search API. Args: objective: Natural language objective describing what you want to find

Parse command line arguments.

Parse the powermetrics output to extract power information. Expected lines in the output: CPU Power: 4382 mW GPU

Pretty print GPU claims dictionary with formatted output. Args: gpu_claims (dict): Dictionary containing GPU attestation claims

Print the attestation analysis. Just for debugging.

Print status of all API keys

Print test summary

Process all pages of a PDF and convert each to markdown. Args: pdf_data: Path to the PDF file or PDF data as bytes prompt: P

Download and register a model on the server.

Query the student model (Ollama).

Read document from file path.

Register a new OAuth2 client.

Render interface to create a blank character for manual editing.

Render the character creation interface.

Render the character editor interface.

Render the character sheet display.

Render the character upload interface.

Render the chat interface with the generated character.

Render the main story generation interface

Render the sidebar with configuration options.

Render the sidebar with configuration options

Rerank documents for given queries using ColBERT's late interaction. Args: queries: List of query strings

Reset character and chat history.

Handle chat completions using the Responses API for reasoning models.

Handle completions using the Groq Responses API. Args: messages: List of message dictionaries with 'role' and 'content'

Retrieve chunks from Qdrant vector database. :param chunks: List of text chunks to embed and search through :param keywords: List of key

Revoke an OAuth2 client.

Revoke an API key.

Run all integration tests.

Run all tests in the suite.

Run all tests in the suite.

Run the full evaluation process and generate a comprehensive report.

Call the NVIDIA Local GPU Verifier (nvtrust) and retrieve the JWT Entity Attestation Token (EAT). Assumes the package pip install nv-

Run hf-mem CLI to estimate inference memory requirements for a HuggingFace model. Uses the hf-mem package (https://github.com/alvarobart

Run the MCP Minion method using only callbacks

Run the SGLang model with the given context.

Run the direct Ollama method (no MCP)

Run the protocol with pre-initialized clients.

Run tests for a specific service

Run all unit tests and display results.

Save comparison results to a JSON file.

Handle synchronous chat completions.

Handle synchronous chat completions. Args: messages: List of message dictionaries with 'role' and 'content' keys

Handle synchronous chat completions with optional MCP tool calling.

Synchronous chat: used for Minion and DeepResearch.

Handle synchronous chat completions.

Handle synchronous chat completions. If you pass a list of message dicts, we do one call for that entire conversation. If you pass a

Scrape a URL using Firecrawl v2 API. Args: url: The URL to scrape api_key: Optional API key. If not provided, reads from

Search documents using the specified retriever and return top k results. Args: query: Search query (string for semantic retrieve

Send JSON-RPC request with streaming response.

Send a task to the A2A server.

Send a task with streaming response.

Send a task with streaming response.

Send task with streaming response.

Set up the AI clients based on user selection

Shutdown the converter and clean up resources.

Split story text into chapters with titles and content. Args: story_text: The full story text Returns: List