plexus.cli.procedure.lua_dsl.runtime module

Lua DSL Runtime - Main execution engine for Lua-based procedures.

Orchestrates: 1. YAML parsing and validation 2. Lua sandbox setup 3. Primitive injection 4. Agent configuration with LLMs and tools 5. Workflow execution

class plexus.cli.procedure.lua_dsl.runtime.LuaDSLRuntime(procedure_id: str, client, mcp_server, openai_api_key: str | None = None)

Bases: object

Main execution engine for Lua-based procedure workflows.

Responsibilities: - Parse and validate YAML configuration - Setup sandboxed Lua environment - Create and inject primitives - Configure agents with LLMs and tools - Execute Lua workflow code - Return results

Initialize the Lua DSL runtime.

Args:

procedure_id: Unique procedure identifier client: PlexusDashboardClient instance mcp_server: MCP server providing tools openai_api_key: Optional OpenAI API key for LLMs

__init__(procedure_id: str, client, mcp_server, openai_api_key: str | None = None)

Initialize the Lua DSL runtime.

Args:

procedure_id: Unique procedure identifier client: PlexusDashboardClient instance mcp_server: MCP server providing tools openai_api_key: Optional OpenAI API key for LLMs

async execute(yaml_config: str, context: Dict[str, Any] | None = None) → Dict[str, Any]

Execute a Lua-based procedure workflow.

Args:

yaml_config: YAML configuration string context: Optional context dict with pre-loaded data

Returns:

Execution results dict with:

• success: bool

• result: Any (return value from Lua workflow)

• state: Final state

• iterations: Number of iterations

• tools_used: List of tool names called

• error: Error message if failed

Raises:

LuaDSLRuntimeError: If execution fails

get_iteration_count() → int

Get current iteration count.

get_state() → Dict[str, Any]

Get current procedure state.

is_stopped() → bool

Check if procedure was stopped.

exception plexus.cli.procedure.lua_dsl.runtime.LuaDSLRuntimeError

Bases: Exception

Raised when Lua DSL runtime encounters an error.