Kimi Agent
b9b78adaa2
Three optimizations to the agentic loop: 1. Cache loop agent as singleton (avoid repeated warmups) 2. Sliding window for step context (last 2 results, not all) 3. Replace summary LLM call with deterministic summary Saves 1 full LLM inference call per agentic loop invocation (30-60s on local models) and reduces context window pressure. Also fixes pre-existing test_cli.py repl test bugs (missing result= assignment).
234 lines
7.4 KiB
Python
234 lines
7.4 KiB
Python
"""Unit tests for the agentic loop module.
|
|
|
|
Tests cover planning, execution, max_steps enforcement, failure
|
|
adaptation, progress callbacks, and response cleaning.
|
|
"""
|
|
|
|
from unittest.mock import AsyncMock, MagicMock, patch
|
|
|
|
import pytest
|
|
|
|
from timmy.agentic_loop import _parse_steps, run_agentic_loop
|
|
|
|
# ---------------------------------------------------------------------------
|
|
# Helpers
|
|
# ---------------------------------------------------------------------------
|
|
|
|
|
|
def _mock_run(content: str):
|
|
"""Create a mock return value for agent.run()."""
|
|
m = MagicMock()
|
|
m.content = content
|
|
return m
|
|
|
|
|
|
# ---------------------------------------------------------------------------
|
|
# _parse_steps
|
|
# ---------------------------------------------------------------------------
|
|
|
|
|
|
class TestParseSteps:
|
|
def test_numbered_with_dot(self):
|
|
text = "1. Search for data\n2. Write to file\n3. Verify"
|
|
assert _parse_steps(text) == ["Search for data", "Write to file", "Verify"]
|
|
|
|
def test_numbered_with_paren(self):
|
|
text = "1) Read config\n2) Update value\n3) Restart"
|
|
assert _parse_steps(text) == ["Read config", "Update value", "Restart"]
|
|
|
|
def test_fallback_plain_lines(self):
|
|
text = "Search the web\nWrite results\nDone"
|
|
assert _parse_steps(text) == ["Search the web", "Write results", "Done"]
|
|
|
|
def test_empty_returns_empty(self):
|
|
assert _parse_steps("") == []
|
|
|
|
|
|
# ---------------------------------------------------------------------------
|
|
# run_agentic_loop
|
|
# ---------------------------------------------------------------------------
|
|
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_planning_phase_produces_steps():
|
|
"""Planning prompt returns numbered step list."""
|
|
mock_agent = MagicMock()
|
|
mock_agent.run = MagicMock(
|
|
side_effect=[
|
|
_mock_run("1. Search AI news\n2. Write to file\n3. Verify"),
|
|
_mock_run("Found 5 articles about AI."),
|
|
_mock_run("Wrote summary to /tmp/ai_news.md"),
|
|
_mock_run("File verified, 15 lines."),
|
|
]
|
|
)
|
|
|
|
with (
|
|
patch("timmy.agentic_loop._get_loop_agent", return_value=mock_agent),
|
|
patch("timmy.agentic_loop._broadcast_progress", new_callable=AsyncMock),
|
|
):
|
|
result = await run_agentic_loop("Search AI news and write summary")
|
|
|
|
assert result.status == "completed"
|
|
assert len(result.steps) == 3
|
|
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_loop_executes_all_steps():
|
|
"""Loop calls agent.run() for plan + each step + summary."""
|
|
mock_agent = MagicMock()
|
|
mock_agent.run = MagicMock(
|
|
side_effect=[
|
|
_mock_run("1. Do A\n2. Do B"),
|
|
_mock_run("A done"),
|
|
_mock_run("B done"),
|
|
]
|
|
)
|
|
|
|
with (
|
|
patch("timmy.agentic_loop._get_loop_agent", return_value=mock_agent),
|
|
patch("timmy.agentic_loop._broadcast_progress", new_callable=AsyncMock),
|
|
):
|
|
result = await run_agentic_loop("Do A and B")
|
|
|
|
# plan + 2 steps = 3 calls
|
|
assert mock_agent.run.call_count == 3
|
|
assert len(result.steps) == 2
|
|
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_loop_respects_max_steps():
|
|
"""Loop stops at max_steps and returns status='partial'."""
|
|
mock_agent = MagicMock()
|
|
mock_agent.run = MagicMock(
|
|
side_effect=[
|
|
_mock_run("1. A\n2. B\n3. C\n4. D\n5. E"),
|
|
_mock_run("A done"),
|
|
_mock_run("B done"),
|
|
]
|
|
)
|
|
|
|
with (
|
|
patch("timmy.agentic_loop._get_loop_agent", return_value=mock_agent),
|
|
patch("timmy.agentic_loop._broadcast_progress", new_callable=AsyncMock),
|
|
):
|
|
result = await run_agentic_loop("Do 5 things", max_steps=2)
|
|
|
|
assert len(result.steps) == 2
|
|
assert result.status == "partial"
|
|
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_failure_triggers_adaptation():
|
|
"""Failed step feeds error back to model, step marked as adapted."""
|
|
mock_agent = MagicMock()
|
|
mock_agent.run = MagicMock(
|
|
side_effect=[
|
|
_mock_run("1. Read config\n2. Update setting\n3. Verify"),
|
|
_mock_run("Config: timeout=30"),
|
|
Exception("Permission denied"),
|
|
_mock_run("Adapted: wrote to ~/config.yaml instead"),
|
|
_mock_run("Verified: timeout=60"),
|
|
]
|
|
)
|
|
|
|
with (
|
|
patch("timmy.agentic_loop._get_loop_agent", return_value=mock_agent),
|
|
patch("timmy.agentic_loop._broadcast_progress", new_callable=AsyncMock),
|
|
):
|
|
result = await run_agentic_loop("Update config timeout to 60")
|
|
|
|
assert result.status == "completed"
|
|
assert any(s.status == "adapted" for s in result.steps)
|
|
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_progress_callback_fires():
|
|
"""on_progress called for each step completion."""
|
|
events = []
|
|
|
|
async def on_progress(desc, step, total):
|
|
events.append((step, total))
|
|
|
|
mock_agent = MagicMock()
|
|
mock_agent.run = MagicMock(
|
|
side_effect=[
|
|
_mock_run("1. Do A\n2. Do B"),
|
|
_mock_run("A done"),
|
|
_mock_run("B done"),
|
|
]
|
|
)
|
|
|
|
with (
|
|
patch("timmy.agentic_loop._get_loop_agent", return_value=mock_agent),
|
|
patch("timmy.agentic_loop._broadcast_progress", new_callable=AsyncMock),
|
|
):
|
|
await run_agentic_loop("Do A and B", on_progress=on_progress)
|
|
|
|
assert len(events) == 2
|
|
assert events[0] == (1, 2)
|
|
assert events[1] == (2, 2)
|
|
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_result_contains_step_metadata():
|
|
"""AgenticResult.steps has status and duration per step."""
|
|
mock_agent = MagicMock()
|
|
mock_agent.run = MagicMock(
|
|
side_effect=[
|
|
_mock_run("1. Search\n2. Write"),
|
|
_mock_run("Found results"),
|
|
_mock_run("Written to file"),
|
|
]
|
|
)
|
|
|
|
with (
|
|
patch("timmy.agentic_loop._get_loop_agent", return_value=mock_agent),
|
|
patch("timmy.agentic_loop._broadcast_progress", new_callable=AsyncMock),
|
|
):
|
|
result = await run_agentic_loop("Search and write")
|
|
|
|
for step in result.steps:
|
|
assert step.status in ("completed", "failed", "adapted")
|
|
assert step.duration_ms >= 0
|
|
assert step.description
|
|
assert step.result
|
|
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_config_default_used():
|
|
"""When max_steps=0, uses settings.max_agent_steps."""
|
|
mock_agent = MagicMock()
|
|
# Return more steps than default config allows (10)
|
|
steps_text = "\n".join(f"{i}. Step {i}" for i in range(1, 15))
|
|
side_effects = [_mock_run(steps_text)]
|
|
# 10 step results
|
|
for i in range(1, 11):
|
|
side_effects.append(_mock_run(f"Step {i} done"))
|
|
|
|
mock_agent.run = MagicMock(side_effect=side_effects)
|
|
|
|
with (
|
|
patch("timmy.agentic_loop._get_loop_agent", return_value=mock_agent),
|
|
patch("timmy.agentic_loop._broadcast_progress", new_callable=AsyncMock),
|
|
):
|
|
result = await run_agentic_loop("Do 14 things", max_steps=0)
|
|
|
|
# Should be capped at 10 (config default)
|
|
assert len(result.steps) == 10
|
|
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_planning_failure_returns_failed():
|
|
"""If the planning phase fails, result.status is 'failed'."""
|
|
mock_agent = MagicMock()
|
|
mock_agent.run = MagicMock(side_effect=Exception("Model offline"))
|
|
|
|
with (
|
|
patch("timmy.agentic_loop._get_loop_agent", return_value=mock_agent),
|
|
patch("timmy.agentic_loop._broadcast_progress", new_callable=AsyncMock),
|
|
):
|
|
result = await run_agentic_loop("Do something")
|
|
|
|
assert result.status == "failed"
|
|
assert "Planning failed" in result.summary
|