RA.Aid/ra_aid/agents/key_facts_gc_agent.py

130 lines
4.5 KiB
Python

"""
Key facts gc agent implementation.
This agent is responsible for maintaining the knowledge base by pruning less important
facts when the total number exceeds a specified threshold. The agent evaluates all
key facts and deletes the least valuable ones to keep the database clean and relevant.
"""
from typing import List
from langchain_core.tools import tool
from rich.console import Console
from rich.markdown import Markdown
from rich.panel import Panel
from ra_aid.agent_utils import create_agent, run_agent_with_retry
from ra_aid.database.repositories.key_fact_repository import KeyFactRepository
from ra_aid.llm import initialize_llm
from ra_aid.prompts.key_facts_gc_prompts import KEY_FACTS_GC_PROMPT
from ra_aid.tools.memory import log_work_event, _global_memory
console = Console()
key_fact_repository = KeyFactRepository()
@tool
def delete_key_facts(fact_ids: List[int]) -> str:
"""Delete multiple key facts by their IDs.
Args:
fact_ids: List of IDs of the key facts to delete
Returns:
str: Success or failure message
"""
deleted_facts = []
not_found_facts = []
failed_facts = []
for fact_id in fact_ids:
# Get the fact first to display information
fact = key_fact_repository.get(fact_id)
if fact:
# Delete the fact
was_deleted = key_fact_repository.delete(fact_id)
if was_deleted:
deleted_facts.append((fact_id, fact.content))
log_work_event(f"Deleted fact {fact_id}.")
else:
failed_facts.append(fact_id)
else:
not_found_facts.append(fact_id)
# Prepare result message
result_parts = []
if deleted_facts:
deleted_msg = "Successfully deleted facts:\n" + "\n".join([f"- #{fact_id}: {content}" for fact_id, content in deleted_facts])
result_parts.append(deleted_msg)
console.print(
Panel(Markdown(deleted_msg), title="Facts Deleted", border_style="green")
)
if not_found_facts:
not_found_msg = f"Facts not found: {', '.join([f'#{fact_id}' for fact_id in not_found_facts])}"
result_parts.append(not_found_msg)
if failed_facts:
failed_msg = f"Failed to delete facts: {', '.join([f'#{fact_id}' for fact_id in failed_facts])}"
result_parts.append(failed_msg)
return "\n".join(result_parts)
def run_key_facts_gc_agent() -> None:
"""Run the key facts gc agent to maintain a reasonable number of key facts.
The agent analyzes all key facts and determines which are the least valuable,
deleting them to maintain a manageable collection size of high-value facts.
"""
# Get the count of key facts
facts = key_fact_repository.get_all()
fact_count = len(facts)
# Display status panel with fact count included
console.print(Panel(f"Gathering my thoughts...\nCurrent number of key facts: {fact_count}", title="🗑 Garbage Collection"))
# Only run the agent if we actually have facts to clean
if fact_count > 0:
# Get all facts as a formatted string for the prompt
facts_dict = key_fact_repository.get_facts_dict()
formatted_facts = "\n".join([f"Fact #{k}: {v}" for k, v in facts_dict.items()])
# Retrieve configuration
llm_config = _global_memory.get("config", {})
# Initialize the LLM model
model = initialize_llm(
llm_config.get("provider", "anthropic"),
llm_config.get("model", "claude-3-7-sonnet-20250219"),
temperature=llm_config.get("temperature")
)
# Create the agent with the delete_key_facts tool
agent = create_agent(model, [delete_key_facts])
# Format the prompt with the current facts
prompt = KEY_FACTS_GC_PROMPT.format(key_facts=formatted_facts)
# Set up the agent configuration
agent_config = {
"recursion_limit": 50 # Set a reasonable recursion limit
}
# Run the agent
run_agent_with_retry(agent, prompt, agent_config)
# Get updated count
updated_facts = key_fact_repository.get_all()
updated_count = len(updated_facts)
# Show info panel with updated count
console.print(
Panel(
f"Cleaned key facts: {fact_count}{updated_count}",
title="🗑 GC Complete"
)
)
else:
console.print(Panel("No key facts to clean.", title="🗑 GC Info"))