Set active provider.

Set tool confirmation mode (legacy compatibility).

Store a provider API key in secure storage. Note: This stores in token storage. Environment variables take precedence and must be set se

Set up environment variables to silence subprocesses before they start.

Show the computed batch structure.

Show final streaming response. Args: content: Final content elapsed: Elapsed time interrupted: Whether interrupted

Visualize execution timing as a timeline.

Convert to dictionary (no API key stored).

Export context for command handlers.

Convert to LLM function calling format (OpenAI/Anthropic compatible).

Unregister a bridge for the given agent_id.

Update an agent's status and broadcast AGENT_STATUS.

Validate provider and print helpful error message if invalid. Args: provider: Provider name to validate Returns

Check if a provider is available.

Validate provider exists, potentially after discovery. Args: provider: Provider name to validate Returns: True if provider e

Wait for an agent to finish. Returns status dict.

Adapt tools for current provider.

Analyze execution traces and generate recommendations.

Build audio content blocks from base64-encoded data.

Build the O(1) tool lookup index from all tools. Index keys: - tool.name → tool (prefers non-default namespaces) - tool.fully

Build tool registry payload from ChatContext's tool_manager.

Clear the current status display without resetting state (async version). This is used when transitioning between display modes (e.g., stream

Convert a raw content block dict into a dashboard-safe descriptor. Used during conversation history replay when we have raw content b

Trim procedural memory patterns to stay within configured limits.

Kick off the full-screen chat UI (banner printed by the handler).

Execute a single plan step using the LLM for tool call generation. When a model_manager is available, the LLM generates the tool call

Execute a step using the plan's static arguments (no LLM).

Execute a step via a full agentic loop. The LLM drives the entire step execution: 1. Sees the step description, tool schemas, and var

Finish the live display and prepare for normal output. This clears the current display and resets state so that subsequent output app

Recursively fix array schemas in place.

Generate a plan dict from a goal description using PlanAgent.

Generate a unique session ID.

Generate a plain text report.

Build image content blocks from base64-encoded data.

Inject internal (non-MCP) tools into the tool list. Idempotent — checks for existing tool names before adding. Called both from _load

Create a mock PreferenceManager.

Create a set of test tools.

List persistent memories.

Record tool execution result in the guard system. Binds the result as a $vN value and increments budget counters.

Render current tool execution status with argument preview.

Try to find the bundled package config as a fallback.

Resolve a dotted variable path like 'api.endpoint.port'.

Execute a plan and return JSON results.

Return the file path for a scope.

Send AGENT_LIST to a specific client.

Setup empty toolset when no servers configured.

Show tools in a table format.

Build text content blocks from a string.

Trigger an immediate refresh (for new content).

Add assistant message to conversation.

Queue a context management notice for the next API call.

Add a system message to conversation history.

Add a pattern-based rule for tool confirmations. Args: pattern: Glob pattern for tool names action: 'always', 'never'

Apply single timeout to all relevant types.

Cancel running tool execution.

Check health of one or all servers. Returns a dict with per-server status: {"server_name": {"status": "healthy"|"unhealthy"|"timeou

Clear the conversation history.

Clear all disabled server preferences.

Clear all validation-disabled tools (for re-validation).

Create a HelpCommand instance.

Configure logging for chuk framework components. Args: suppress: If True, suppress INFO/DEBUG logs. If False, allow all.

Configure a provider with API settings. Args: provider: Provider name api_key: API key (optional)

Prompt user to confirm tool execution. Args: tool_name: Name of tool to execute arguments: Tool arguments Re

Convert list of message dicts to Message models. Uses chuk_llm's Message.model_validate for proper Pydantic parsing. Args: messages:

Create a prompt session with transparent autocomplete.

Create a chuk-term table for tools (does not print it).

Create a new Typer app with all commands registered. Returns: Configured Typer application.

Enable a server in preferences.

Execute the tool command.

Execute the help command.

Execute a single tool call with semaphore control.

Find a server by its name.

Clean up after all tool calls complete.

Force a fresh discovery (useful for debugging). Returns: Number of new functions discovered

Create from dictionary.

Create from dictionary format with environment variable handling. Uses ServerConfigInput Pydantic model for validation instead of manual .get

Convert a raw list item (dict | str | int | …) into a ResourceInfo. If *raw* is not a mapping we treat it as an opaque scalar and st

Create from legacy tuple format.

Get all per-tool confirmation settings.

Retrieve an artifact by ID, or None.

Get the bridge for a running app by tool name.

Get detailed help for a specific command. Args: command_name: Name of the command. Returns: Detaile

Get completions for slash commands.

Get a summary of the configuration for debugging.