Created a test of MemoryAgent with thought extraction

tests/test_agents.py

@@ -3,13 +3,18 @@
 import itertools
 import json
 import re
+import time
 from pathlib import Path
 from typing import Any, cast
 from unittest.mock import patch
 
+import ldp.agent
 import pytest
-from aviary.tools import ToolsAdapter, ToolSelector
+from aviary.tools import ToolRequestMessage, ToolsAdapter, ToolSelector
 from ldp.agent import SimpleAgent
+from ldp.graph.memory import Memory, UIndexMemoryModel
+from ldp.graph.ops import OpResult
+from ldp.llms import EmbeddingModel
 from pydantic import ValidationError
 from pytest_subtests import SubTests
 
@@ -26,7 +31,7 @@
 from paperqa.docs import Docs
 from paperqa.settings import AgentSettings, Settings
 from paperqa.types import Answer, Context, Doc, Text
-from paperqa.utils import get_year, md5sum
+from paperqa.utils import extract_thought, get_year, md5sum
 
 
 @pytest.mark.asyncio
@@ -103,6 +108,86 @@ async def test_agent_types(
         assert response.answer.cost > 0, "Expected nonzero cost"
 
 
+@pytest.mark.asyncio
+async def test_successful_memory_agent() -> None:
+    tic = time.perf_counter()
+    memory_id = "call_Wtmv95JbNcQ2nRQCZBoOfcJy"  # Stub value
+    memory = Memory(
+        query=(
+            "Use the tools to answer the question: Q: Acinetobacter lwoffii has"
+            " been evolved in the lab to be resistant to which of these"
+            " antibiotics?\n\nOptions:\nA) gentamicin\nB) Insufficient information"
+            " to answer this question\nC) meropenem\nD) ciproflaxin\nE)"
+            " ampicillin\n\nThe gen_answer tool output is visible to the user, so"
+            " you do not need to restate the answer and can simply terminate if"
+            " the answer looks sufficient. The current status of"
+            " evidence/papers/cost is Status: Paper Count=0 | Relevant Papers=0 |"
+            " Current Evidence=0 | Current Cost=$0.00\n\nTool request message ''"
+            f" for tool call: {memory_id} with content"
+            " 'paper_search(query='Acinetobacter lwoffii antibiotic resistance"
+            " evolution', min_year='None', max_year='None')'\n\nTool response"
+            " message 'Status: Paper Count=11 | Relevant Papers=0 | Current"
+            f" Evidence=0 | Current Cost=$0.00' for tool call ID {memory_id} of"
+            " tool 'paper_search'"
+        ),
+        input=(
+            "Use the tools to answer the question: Q: Acinetobacter lwoffii has"
+            " been evolved in the lab to be resistant to which of these"
+            " antibiotics?\n\nOptions:\nA) gentamicin\nB) Insufficient information"
+            " to answer this question\nC) meropenem\nD) ciproflaxin\nE)"
+            " ampicillin\n\nThe gen_answer tool output is visible to the user, so"
+            " you do not need to restate the answer and can simply terminate if the"
+            " answer looks sufficient. The current status of evidence/papers/cost"
+            " is Status: Paper Count=0 | Relevant Papers=0 | Current Evidence=0 |"
+            " Current Cost=$0.00"
+        ),
+        output=(
+            f"Tool request message '' for tool call: {memory_id} with content"
+            " 'paper_search(query='Acinetobacter lwoffii"
+            " antibiotic resistance evolution', min_year='None',"
+            " max_year='None')'"
+        ),
+        value=5.61,
+        template="Input: {input}\n\nOutput: {output}\n\nDiscounted Reward: {value}",
+    )
+    memory_model = UIndexMemoryModel(
+        embedding_model=EmbeddingModel.from_name("text-embedding-3-small")
+    )
+    await memory_model.add_memory(memory)
+    serialized_memory_model = memory_model.model_dump(exclude_none=True)
+    query = QueryRequest(
+        query=(
+            "Use the tools to answer the question: Q: Acinetobacter lwoffii has"
+            " been evolved in the lab to be resistant to which of these"
+            " antibiotics?\n\nOptions:\nA) gentamicin\nB) Insufficient information"
+            " to answer this question\nC) meropenem\nD) ciproflaxin\nE) ampicillin"
+        )
+    )
+    # NOTE: use Claude 3 for its <thinking> feature, testing regex replacement of it
+    query.settings.agent.agent_llm = "claude-3-opus-20240229"
+    query.settings.agent.agent_config = {
+        "memories": serialized_memory_model.pop("memories"),
+        "memory_model": serialized_memory_model,
+    }
+
+    thoughts: list[str] = []
+
+    async def on_agent_action(action: OpResult[ToolRequestMessage], *_) -> None:
+        thoughts.append(extract_thought(content=action.value.content))
+
+    response = await agent_query(
+        query,
+        Docs(),
+        agent_type=f"{ldp.agent.__name__}.MemoryAgent",
+        on_agent_action_callback=on_agent_action,
+    )
+    assert response.status == AgentStatus.SUCCESS, "Agent did not succeed"
+    assert (
+        time.perf_counter() - tic <= query.settings.agent.timeout
+    ), "Agent should not have timed out"
+    assert all("<thinking>" not in thought for thought in thoughts)
+
+
 @pytest.mark.parametrize("agent_type", [ToolSelector, SimpleAgent])
 @pytest.mark.asyncio
 async def test_timeout(agent_test_settings: Settings, agent_type: str | type) -> None: