Files
Timmy-time-dashboard/tests/test_agent.py

196 lines
6.8 KiB
Python
Raw Normal View History

from unittest.mock import MagicMock, patch
def test_create_timmy_returns_agent():
"""create_timmy should delegate to Agno Agent with correct config."""
with patch("timmy.agent.Agent") as MockAgent, \
patch("timmy.agent.Ollama"), \
patch("timmy.agent.SqliteDb"):
mock_instance = MagicMock()
MockAgent.return_value = mock_instance
from timmy.agent import create_timmy
result = create_timmy()
assert result is mock_instance
MockAgent.assert_called_once()
def test_create_timmy_agent_name():
with patch("timmy.agent.Agent") as MockAgent, \
patch("timmy.agent.Ollama"), \
patch("timmy.agent.SqliteDb"):
from timmy.agent import create_timmy
create_timmy()
kwargs = MockAgent.call_args.kwargs
assert kwargs["name"] == "Timmy"
def test_create_timmy_uses_llama32():
with patch("timmy.agent.Agent"), \
patch("timmy.agent.Ollama") as MockOllama, \
patch("timmy.agent.SqliteDb"):
from timmy.agent import create_timmy
create_timmy()
MockOllama.assert_called_once()
kwargs = MockOllama.call_args.kwargs
assert kwargs["id"] == "llama3.2"
def test_create_timmy_history_config():
with patch("timmy.agent.Agent") as MockAgent, \
patch("timmy.agent.Ollama"), \
patch("timmy.agent.SqliteDb"):
from timmy.agent import create_timmy
create_timmy()
kwargs = MockAgent.call_args.kwargs
assert kwargs["add_history_to_context"] is True
Fine-tune Timmy's conversational AI with memory layers ## Enhanced System Prompt - Detailed tool usage guidelines with explicit examples - Clear DO and DON'T examples for tool selection - Memory system documentation - Conversation flow guidelines - Context awareness instructions ## Memory Layer System (NEW) Implemented 3-layer memory architecture: 1. WORKING MEMORY (src/timmy/memory_layers.py) - Immediate context (last 20 messages) - Topic tracking - Tool call tracking - Fast, ephemeral 2. SHORT-TERM MEMORY (Agno SQLite) - Recent conversations (100) - Persists across restarts - Managed by Agno Agent 3. LONG-TERM MEMORY (src/timmy/memory_layers.py) - Facts about user (name, preferences) - SQLite storage in data/memory/ - Auto-extraction from conversations - User profile generation ## Memory Manager (NEW) - Central coordinator for all memory layers - Context injection into prompts - Fact extraction and storage - Session management ## TimmyWithMemory Class (NEW) - Wrapper around Agno Agent with explicit memory - Auto-injects user context from LTM - Tracks exchanges across all layers - Simple chat() interface ## Agent Configuration - Increased num_history_runs: 10 -> 20 - Better conversational context retention ## Tests - All 973 tests pass - Fixed test expectations for new config - Fixed module path in test_scary_paths.py ## Files Added/Modified - src/timmy/prompts.py - Enhanced with memory and tool guidance - src/timmy/agent.py - Added TimmyWithMemory class - src/timmy/memory_layers.py - NEW memory system - src/timmy/conversation.py - NEW conversation manager - tests/ - Updated for new config
2026-02-25 18:07:44 -05:00
assert kwargs["num_history_runs"] == 20
assert kwargs["markdown"] is True
def test_create_timmy_custom_db_file():
with patch("timmy.agent.Agent"), \
patch("timmy.agent.Ollama"), \
patch("timmy.agent.SqliteDb") as MockDb:
from timmy.agent import create_timmy
create_timmy(db_file="custom.db")
MockDb.assert_called_once_with(db_file="custom.db")
def test_create_timmy_embeds_system_prompt():
from timmy.prompts import TIMMY_SYSTEM_PROMPT
with patch("timmy.agent.Agent") as MockAgent, \
patch("timmy.agent.Ollama"), \
patch("timmy.agent.SqliteDb"):
from timmy.agent import create_timmy
create_timmy()
kwargs = MockAgent.call_args.kwargs
assert kwargs["description"] == TIMMY_SYSTEM_PROMPT
# ── Ollama host regression (container connectivity) ─────────────────────────
def test_create_timmy_passes_ollama_url_to_model():
"""Regression: Ollama model must receive settings.ollama_url as host.
Without this, containers default to localhost:11434 which is unreachable
when Ollama runs on the Docker host.
"""
with patch("timmy.agent.Agent"), \
patch("timmy.agent.Ollama") as MockOllama, \
patch("timmy.agent.SqliteDb"):
from timmy.agent import create_timmy
create_timmy()
kwargs = MockOllama.call_args.kwargs
assert "host" in kwargs, "Ollama() must receive host= parameter"
assert kwargs["host"] == "http://localhost:11434" # default from config
def test_create_timmy_respects_custom_ollama_url():
"""Ollama host should follow OLLAMA_URL when overridden in config."""
custom_url = "http://host.docker.internal:11434"
with patch("timmy.agent.Agent"), \
patch("timmy.agent.Ollama") as MockOllama, \
patch("timmy.agent.SqliteDb"), \
patch("timmy.agent.settings") as mock_settings:
mock_settings.ollama_model = "llama3.2"
mock_settings.ollama_url = custom_url
mock_settings.timmy_model_backend = "ollama"
mock_settings.airllm_model_size = "70b"
from timmy.agent import create_timmy
create_timmy()
kwargs = MockOllama.call_args.kwargs
assert kwargs["host"] == custom_url
# ── AirLLM path ──────────────────────────────────────────────────────────────
def test_create_timmy_airllm_returns_airllm_agent():
"""backend='airllm' must return a TimmyAirLLMAgent, not an Agno Agent."""
with patch("timmy.backends.is_apple_silicon", return_value=False):
from timmy.agent import create_timmy
from timmy.backends import TimmyAirLLMAgent
result = create_timmy(backend="airllm", model_size="8b")
assert isinstance(result, TimmyAirLLMAgent)
def test_create_timmy_airllm_does_not_call_agno_agent():
"""When using the airllm backend, Agno Agent should never be instantiated."""
with patch("timmy.agent.Agent") as MockAgent, \
patch("timmy.backends.is_apple_silicon", return_value=False):
from timmy.agent import create_timmy
create_timmy(backend="airllm", model_size="8b")
MockAgent.assert_not_called()
def test_create_timmy_explicit_ollama_ignores_autodetect():
"""backend='ollama' must always use Ollama, even on Apple Silicon."""
with patch("timmy.agent.Agent") as MockAgent, \
patch("timmy.agent.Ollama"), \
patch("timmy.agent.SqliteDb"):
from timmy.agent import create_timmy
create_timmy(backend="ollama")
MockAgent.assert_called_once()
# ── _resolve_backend ─────────────────────────────────────────────────────────
def test_resolve_backend_explicit_takes_priority():
from timmy.agent import _resolve_backend
assert _resolve_backend("airllm") == "airllm"
assert _resolve_backend("ollama") == "ollama"
def test_resolve_backend_defaults_to_ollama_without_config():
"""Default config (timmy_model_backend='ollama') → 'ollama'."""
from timmy.agent import _resolve_backend
assert _resolve_backend(None) == "ollama"
def test_resolve_backend_auto_uses_airllm_on_apple_silicon():
"""'auto' on Apple Silicon with airllm stubbed → 'airllm'."""
with patch("timmy.backends.is_apple_silicon", return_value=True), \
patch("timmy.agent.settings") as mock_settings:
mock_settings.timmy_model_backend = "auto"
mock_settings.airllm_model_size = "70b"
mock_settings.ollama_model = "llama3.2"
from timmy.agent import _resolve_backend
assert _resolve_backend(None) == "airllm"
def test_resolve_backend_auto_falls_back_on_non_apple():
"""'auto' on non-Apple Silicon → 'ollama'."""
with patch("timmy.backends.is_apple_silicon", return_value=False), \
patch("timmy.agent.settings") as mock_settings:
mock_settings.timmy_model_backend = "auto"
mock_settings.airllm_model_size = "70b"
mock_settings.ollama_model = "llama3.2"
from timmy.agent import _resolve_backend
assert _resolve_backend(None) == "ollama"