Print a human-readable summary of metrics

Read data from Excel file. Args: file_path: Path to Excel file sheet_name: Specific sheet name (None for all sheets)

Read a file and return its contents. Args: file_path: Path to the file encoding: File encoding (default: utf-8)

Read multiple files at once. Args: file_paths: List of file paths encoding: File encoding

Read and extract content from a webpage. Args: url: URL of the webpage extract_text: Whether to extract main text conten

Register a new tool

Remove descriptions from tools and their parameters (including nested structures) Args: tools_info: Original tools information

Replay a workflow with the given parameters. Args: workflow: The workflow to replay parameters: Para

Request approval from a human administrator. Args: request_message: Message describing what needs approval context: Optional

Rerank documents for a query. Args: query: The search query documents: List of documents with text and metada

* Reset VAD state

* Reset Silero VAD state tensors

Reset the board to the starting position. Returns: Dictionary with reset board state

Clear conversation history and current content

Main CLI loop

Run all test cases

Run a single ablation experiment Returns: Tuple of (experiment_name, success_rate)

Run experiment with traditional Q-learning agent. Args: num_training_episodes: Number of episodes to train

Run the API server.

Run a single test case Args: test_name: Name of test to run

Sanitize a single conversation and collect metrics Returns: Dictionary with sanitized conversation and metrics

Save the Q-table and parameters.

Save experiences to file for analysis.

Save conversation history to storage

Save the RAPTOR tree index to disk.

Save the knowledge graph index to disk.

Save all metrics and summary to files

Save sanitized logs to output directory

Calculate total BM25 score for a document given query terms

Search the RAPTOR tree for relevant information.

Search for documents.

Search for documents.

Search for similar documents.

Search memory cards by query. Args: query: Search query Returns: List of (memor

Search user memory for relevant information Args: query: Natural language search query top_k: Number

Search user memory for relevant information Args: query: Natural language search query top_k: Number

Search the web using DuckDuckGo (free, no API key required). Args: query: The search query string num_results: Number of

Search Wikipedia and get article summary. Args: query: Search query language: Wikipedia language (en, zh, etc.)

Interactively prompt the user to select a memory mode Returns: Selected MemoryMode

Interactive mode selection menu Returns: Selected mode string or None for all modes

Send an event to the agent Args: event_type: Type of event (e.g., 'web_message', 'im_message') conte

Set a timer that will notify when completed. Args: duration_seconds: How long to wait before timer expires timer_name: Option

Initialize Playwright and browser.

Start the shell process

Start system monitoring

Summarize a successful trajectory into reusable experience. Args: question: The original question re

Summarize complex tool output. Args: tool_name: Name of the tool that produced the output output: Th

Summarize text using OpenAI API.

Convert to dictionary for JSON serialization

* Transcribe audio using the configured ASR provider * @param {Buffer} audioBuffer - Raw audio data * @param {Object} options - Transcription op

Store attention weights for a position (only for output tokens)

Store an experience for future in-context learning.

Update an existing memory item

验证配置是否有效 Returns: bool: 配置是否有效

Check which API keys are configured

Initialize the game environment. Args: seed: Random seed for reproducibility stochastic: If True, ad

Advance simulated time for demo purposes

Advance simulated time for demo purposes

Analyze types of actions in trajectory. Args: trajectory: Execution trajectory Returns:

Analyze comparison results to highlight improvements

Use Doubao for content analysis

Use Gemini for audio analysis with thinking mode

Use Gemini for PDF analysis with thinking mode

Use OpenAI for content analysis

Attack with a weapon.

Build BM25 indexes for both contextual and non-contextual chunks. Educational Note: BM25 uses TF-IDF with optimizati

Build vLLM server command with arguments

Build context based on the current mode Returns: Context string for the model

Build the evaluation prompt for the LLM.

Build full conversation text from chunks

Build messages for the LLM including conversation history

Build messages for the LLM including conversation history

Build the OpenRouter-specific request format matching the Go implementation Args: messages: Conversation message

Build the full prompt for the LLM.

Build enhanced system prompt with memory cards

Calculate how well a workflow matches a task description. Args: task: Normalized task description wo

Calculate statistics for educational purposes. Args: dense_results: Dense search results sparse_results: Spar

Call the LLM with retry logic.

Capture the current step for workflow learning.

调用 Kimi API 进行对话 Args: messages: 消息列表 Returns: API 响应的 Choice 对象

Stream chat response (only when tools are disabled) Args: message: User message Returns:

Check for lint errors

Check for lint errors after file modification

Check for lint errors after file modification

Check if the retrieval pipeline service is available

Check if the retrieval pipeline service is available