"""
Single agent chat CLI interface
"""
import sys
from src.config import config
from src.cli.utils import get_or_create_cli_agent
def chat_cli():
"""CLI interface for single agent chat"""
print("š¤ Single Agent Chat")
print("=" * 30)
# Get parameters from command line or interactively
model_id = None
provider = None
role = "Assistant"
goal = "Help users with their requests"
backstory = "A helpful AI assistant"
agent_id = None
session_id = None
enable_memory = True
user_id = "cli_user"
# Sampling parameters
temperature = None
top_p = None
top_k = None
max_tokens = None
presence_penalty = None
frequency_penalty = None
stop = None
seed = None
logprobs = None
reasoning_effort = None
# Parse command line arguments
args = sys.argv[2:] # Skip 'main.py chat'
query = None
stream = False
i = 0
while i < len(args):
if args[i] == "--model" and i + 1 < len(args):
model_id = args[i + 1]
i += 2
elif args[i] == "--provider" and i + 1 < len(args):
provider = args[i + 1]
i += 2
elif args[i] == "--role" and i + 1 < len(args):
role = args[i + 1]
i += 2
elif args[i] == "--goal" and i + 1 < len(args):
goal = args[i + 1]
i += 2
elif args[i] == "--backstory" and i + 1 < len(args):
backstory = args[i + 1]
i += 2
elif args[i] == "--agent-id" and i + 1 < len(args):
agent_id = args[i + 1]
i += 2
elif args[i] == "--session-id" and i + 1 < len(args):
session_id = args[i + 1]
i += 2
elif args[i] == "--user-id" and i + 1 < len(args):
user_id = args[i + 1]
i += 2
elif args[i] == "--no-memory":
enable_memory = False
i += 1
elif args[i] == "--stream":
stream = True
i += 1
# Sampling parameters
elif args[i] == "--temperature" and i + 1 < len(args):
temperature = float(args[i + 1])
i += 2
elif args[i] == "--top-p" and i + 1 < len(args):
top_p = float(args[i + 1])
i += 2
elif args[i] == "--top-k" and i + 1 < len(args):
top_k = int(args[i + 1])
i += 2
elif args[i] == "--max-tokens" and i + 1 < len(args):
max_tokens = int(args[i + 1])
i += 2
elif args[i] == "--presence-penalty" and i + 1 < len(args):
presence_penalty = float(args[i + 1])
i += 2
elif args[i] == "--frequency-penalty" and i + 1 < len(args):
frequency_penalty = float(args[i + 1])
i += 2
elif args[i] == "--stop" and i + 1 < len(args):
stop = args[i + 1]
i += 2
elif args[i] == "--seed" and i + 1 < len(args):
seed = int(args[i + 1])
i += 2
elif args[i] == "--logprobs" and i + 1 < len(args):
logprobs = int(args[i + 1])
i += 2
elif args[i] == "--reasoning-effort" and i + 1 < len(args):
reasoning_effort = args[i + 1]
i += 2
else:
# Assume it's the query
query = " ".join(args[i:])
break
# Handle predefined agent
if agent_id:
predefined_agents = config.get_predefined_agents()
if agent_id in predefined_agents:
agent_config = predefined_agents[agent_id]
model_id = agent_config.get("model_id", model_id)
provider = agent_config.get("provider", provider)
role = agent_config.get("role", role)
goal = agent_config.get("goal", goal)
backstory = agent_config.get("backstory", backstory)
enable_memory = agent_config.get("enable_memory", enable_memory)
else:
print(f"ā Predefined agent '{agent_id}' not found")
return
# Set defaults if not provided
if not model_id:
model_id = str(config.get('models.default', 'gpt-3.5-turbo'))
if not provider:
provider = "openai" # Default provider
if not query:
print("Usage: python main.py chat [options] \"your query\"")
print("\nOptions:")
print(" --model MODEL_ID Model to use (default: gpt-3.5-turbo)")
print(" --provider PROVIDER Provider (default: openai)")
print(" --agent-id AGENT_ID Use predefined agent configuration")
print(" --role ROLE Agent role")
print(" --goal GOAL Agent goal")
print(" --backstory BACKSTORY Agent backstory")
print(" --session-id SESSION_ID Session identifier for memory")
print(" --user-id USER_ID User identifier (default: cli_user)")
print(" --no-memory Disable memory for the agent")
print(" --stream Enable streaming response")
print("\nSampling Parameters:")
print(" --temperature TEMP Controls randomness (0.0-2.0)")
print(" --top-p P Nucleus sampling (0.0-1.0)")
print(" --top-k K Top-k sampling (1-1000)")
print(" --max-tokens TOKENS Maximum tokens to generate")
print(" --presence-penalty P Presence penalty (-2.0-2.0)")
print(" --frequency-penalty P Frequency penalty (-2.0-2.0)")
print(" --stop WORD Stop sequence")
print(" --seed SEED Random seed for reproducibility")
print(" --logprobs N Return log probabilities")
print(" --reasoning-effort EFFORT Reasoning effort (low/medium/high)")
print("\nExamples:")
print(" python main.py chat --model claude-3-haiku-20240307 --provider anthropic \"Hello, how are you?\"")
print(" python main.py chat --stream --temperature 0.7 --max-tokens 100 \"Tell me a story\"")
print(" python main.py chat --agent-id financial_analyst \"Analyze this market trend\"")
return
try:
# Create agent with full sampling parameters
print(f"Creating agent with model: {model_id} ({provider})")
if agent_id:
print(f"Using predefined agent: {agent_id}")
agent = get_or_create_cli_agent(
model_id=model_id,
provider=provider,
role=role,
goal=goal,
backstory=backstory,
enable_memory=enable_memory,
session_id=session_id,
temperature=temperature,
top_p=top_p,
top_k=top_k,
max_tokens=max_tokens,
presence_penalty=presence_penalty,
frequency_penalty=frequency_penalty,
stop=stop,
seed=seed,
logprobs=logprobs,
reasoning_effort=reasoning_effort,
)
# Process query
print(f"\nš Query: {query}")
if stream:
print("š¤ Response (streaming): ", end="", flush=True)
else:
print("š¤ Response: ", end="", flush=True)
response = agent.process_request(query, stream=stream)
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"ā Error: {e}")
