"""
Multi-agent coordination CLI interface
"""
import sys
from src.config import config
from src.agents import MultiAgentCoordinator, CoordinationMode
from src.cli.utils import get_or_create_cli_agent
def multi_agent_cli():
"""CLI interface for multi-agent coordination"""
print("👥 Multi-Agent Coordination")
print("=" * 35)
# Parse command line arguments
args = sys.argv[2:] # Skip 'main.py multi-agent'
query = None
mode = "ensemble"
rebuttal_limit = 3
agent_ids = []
stream = False
i = 0
while i < len(args):
if args[i] == "--mode" and i + 1 < len(args):
mode = args[i + 1]
i += 2
elif args[i] == "--rebuttal-limit" and i + 1 < len(args):
rebuttal_limit = int(args[i + 1])
i += 2
elif args[i] == "--agents" and i + 1 < len(args):
agent_ids = args[i + 1].split(",")
i += 2
elif args[i] == "--stream":
stream = True
i += 1
else:
# Assume it's the query
query = " ".join(args[i:])
break
if not query:
print("Usage: python main.py multi-agent [options] \"your query\"")
print("\nOptions:")
print(" --mode MODE Coordination mode: pipeline, ensemble, debate, swarm, hierarchical (default: ensemble)")
print(" --rebuttal-limit NUM Max rebuttals for debate mode (default: 3)")
print(" --agents ID1,ID2,... Comma-separated agent IDs (default: financial_analyst,strategy_consultant)")
print(" --stream Enable streaming response")
print("\nAvailable agents:")
agents_data = config.get_predefined_agents()
for agent_id in agents_data.keys():
print(f" - {agent_id}")
print("\nExamples:")
print(" python main.py multi-agent --mode debate --agents financial_analyst,strategy_consultant \"Should companies prioritize profit or sustainability?\"")
print(" python main.py multi-agent --stream --mode ensemble --agents technical_expert,data_scientist \"Analyze this dataset\"")
return
# Set default agents if not specified
if not agent_ids:
agents_data = config.get_predefined_agents()
agent_ids = list(agents_data.keys())[:2] # Use first 2 agents
try:
# Get predefined agents
agents_data = config.get_predefined_agents()
agents = []
for agent_id in agent_ids:
if agent_id not in agents_data:
print(f"❌ Agent '{agent_id}' not found")
return
agent_config = agents_data[agent_id]
agent = get_or_create_cli_agent(
model_id=str(agent_config.get("model_id", "gpt-3.5-turbo")),
provider=str(agent_config.get("provider", "openai")),
role=str(agent_config.get("role", "Assistant")),
goal=str(agent_config.get("goal", "Help with tasks")),
backstory=str(agent_config.get("backstory", "An AI assistant")),
enable_memory=agent_config.get("enable_memory", True),
session_id=agent_config.get("session_id"),
)
agents.append(agent)
if len(agents) < 2:
print("❌ Need at least 2 agents for multi-agent coordination")
return
# Create coordinator
coordinator = MultiAgentCoordinator(agents)
# Map mode string to enum
mode_map = {
"pipeline": CoordinationMode.PIPELINE,
"ensemble": CoordinationMode.ENSEMBLE,
"debate": CoordinationMode.DEBATE,
"swarm": CoordinationMode.SWARM,
"hierarchical": CoordinationMode.HIERARCHICAL
}
if mode not in mode_map:
print(f"❌ Unknown mode: {mode}")
print(f"Available modes: {', '.join(mode_map.keys())}")
return
coord_mode = mode_map[mode]
# Check for streaming option
stream = "--stream" in args
# Process query
print(f"🤝 Mode: {mode}")
print(f"👥 Agents: {', '.join(agent_ids)}")
print(f"💭 Query: {query}")
if stream:
print("🤖 Response (streaming): ", end="", flush=True)
else:
print("🤖 Response: ", end="", flush=True)
kwargs = {}
if mode == "debate":
kwargs["rebuttal_limit"] = rebuttal_limit
response = coordinator.coordinate(coord_mode, query, stream=stream, **kwargs)
if stream:
# Handle streaming response
for chunk in response:
print(chunk, end="", flush=True)
print() # New line after streaming
else:
print(response)
except Exception as e:
print(f"❌ Error: {e}")
