Save the development session log.

Encrypt and authenticate a single chunk of data.

Configure LMCache environment variables.

Set up FastAPI routes.

Set up graceful shutdown handlers.

Handle single message chat using the /chat endpoint.

Start background task for periodic cleanup.

Starts the `server_name` MCP server via the command specified in the config manager

Start background task for periodic metrics updates.

Generate a summary of the given turns and combine with existing summary. Args: turns_to_summarize: The conversat

Upload content of content_type to specified s3 bucket. :param content: media body to be uploaded :param content_type: type o

Validate that the supervisor URL uses HTTPS protocol for security

Validate and warn about reasoning_effort compatibility with the model. Different GPT-5 model variants support different reas

Validate that the supervisor URL uses HTTPS protocol for security

Validate that the supervisor URL uses HTTPS protocol for security

Validate if the current model supports URL context feature. Returns: bool: True if model supports URL context, F

Wrapper for async chat. Runs `asyncio.run()` internally to simplify usage.

Wrapper for async chat. Runs `asyncio.run()` internally to simplify usage. Args: messages: List of message dictionaries

Wrapper for async chat. Runs `asyncio.run()` internally to simplify usage.

Parallel asynchronous chat for Lemonade. Accepts a list of message dicts or a single dict. This is only used for the Minions

Wrapper for async chat. Runs `asyncio.run()` internally to simplify usage.

Wrapper for async chat. Runs `asyncio.run()` internally to simplify usage.

Analyze and summarize the security properties from an attestation payload. Args: payload: The verified JWT payload from verify_a

Have GPT-4o analyze the task to determine required skills.

Answer a question using Ollama client with concatenated documents as context using RAG. Args: user_query: The user's question

Run in batch mode with predefined queries.

Run batch test of conversation summarization.

Route multiple queries in batch. Args: queries: List of queries to route available_clients: Same as in route

Retrieves top k chunks using BM25 with weighted keywords.

Use GPT-4o to expand a character description into a full exposé. Args: description: Brief character description openai_a

Cache a character exposé to avoid regeneration.

Cancel a task.

Cancel a running task.

Handle chat completions, routing to async or sync implementation. The client will automatically detect URLs in the message c

Handle chat completions using the Distributed Inference API. Args: messages: List of message dictionaries with 'role' an

Handle chat completions using the Cartesia MLX client. Args: messages: List of message dictionaries with 'role' and 'con

Handle chat completions using the Groq API via OpenAI compatibility. Args: messages: List of message dictionaries with '

Handle chat completions using the Transformers model. For Hunyuan models with enable_thinking=True, responses are automatica

Handle chat completions using the Grok API. When files are attached (via file_ids or add_file()), the document_search tool i

Handle chat completions using the Ollama Turbo API. Args: messages: List of message dictionaries with 'role' and

Handle chat completions using either native Cohere SDK or OpenAI compatibility API. Args: messages: List of message dict

Handle chat completions using the Together API.

Handle chat completions using the DeepSeek API. Args: messages: List of message dictionaries with 'role' and 'content' k

Handle chat completions using the secure endpoint. Args: messages: List of message dictionaries with 'role' and 'content

Process chat messages using SGLang.

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

Example of using the chat API.

Example of using the chat API.

Run an interactive chat loop with the protocol.

Chat with the character using Ollama streaming. Args: message: User message character_expose: Character exposé dictionar

Check if Ollama model exists and pull it if not.

Check if required dependencies are available.

Check if required dependencies are installed.

Check if required environment variables are set.

Check if all servers are healthy.

If we are using a local client, we want to be able to check if the local server is running or not

Check if all prerequisites are met for running the app.

Check if the Osaurus server is running and reachable. Returns: Dict: Server response (typically model list)

Helper to group sentences into chunks with overlap.

Clean and format story text for better presentation. Args: story_text: Raw story text Returns: Cleaned and

Clean up test directories.

Clear the client pool (useful for testing or cleanup).

Click on a UI element in an application by description or coordinates.

Close the client.

Close the test client.

Retrieves top k chunks using Liquid AI's ColBERT late-interaction retriever. This function uses PyLate's PLAID index for efficient simil

Example of using the completion API.

Convert chat history to CSV format for download.

Create a backup of the current workspace state. Args: label: Optional label for the backup

Create a blank character template for editing. Returns: Dictionary with empty character template

Create a documentation file for a development step. Args: step_number: The step number content: The

Create a protocol instance based on configuration.

Create a list of sample tasks for testing.

Verify & decode a NVIDIA GPU-EAT that may lack `kid`. Returns the claim set or raises jwt.InvalidSignatureError.

Decrypt and verify a message using sequential processing. This function has the same API as decrypt_and_verify but processes large messages

Delete the current collection from Qdrant. :returns Whether the collection was successfully deleted

Display metrics in a formatted way

Display the status of prerequisites in the sidebar.

Display sample tasks in the sidebar.

Display the generated story with pinwheel navigation

Generate embeddings for the given text content. Args: content: Text string or list of text strings to embed

Embed content using model (must support embeddings).

Generate embeddings using Cohere's embedding API. Args: content: Text content to embed (single string or list of