arcade-mcp/toolkits/search/evals/eval_google_search.py
Sam Partee b6b4cd0a4c
🏗️ Restructure: Multi-Package Architecture + uv Migration (#412)
### Overview
Major restructuring from monolithic `arcade-ai` package to modular
library architecture with standardized uv-based dependency management.

![arcade-ai Monorepo
(2)](https://github.com/user-attachments/assets/25f102b0-bb87-4a04-9701-d227d05664b1)

### New Package Structure
- **`arcade-tdk`** - Lightweight toolkit development kit (core
decorators, auth)
- **`arcade-core`** - Core execution engine and catalog functionality  
- **`arcade-serve`** - FastAPI/MCP server components
- **`arcade-ai`** - Meta package that includes CLI functionality.
Optionally include evals via the `evals` extra. Optionally include all
packages via the `all` extra.

### Key Benefits
- **Lighter Dependencies**: Toolkits now depend only on `arcade-tdk` (~2
deps) vs full `arcade-ai` (~30+ deps)
- **Faster Builds**: uv provides 10-100x faster dependency resolution
and installation
- **Better Modularity**: Clear separation of concerns, consumers import
only what they need
- **Standard Tooling**: Eliminates custom poetry scripts, uses standard
Python packaging

### Migration Impact
- All 20 toolkits converted from poetry → uv with `arcade-tdk`
dependencies plus `arcade-ai[evals]` and `arcade-serve` dev
dependencies. When developing locally, devs should install toolkits via
`make install-local`.
- Modern Python 3.10+ type hints throughout
- Standardized build system with hatchling backend
- Enhanced Makefile with robust toolkit management commands
- Removed `arcade dev` CLI command
- Reduce the number of files created by `arcade new` and add an option
to not generate a tests and evals folder.

This foundation enables faster development cycles and cleaner dependency
chains for the growing toolkit ecosystem.

### Todo After this PR is merged
- [ ] Post-merge workflow(s) (release & publish containers, etc)
- [ ] Release order plan. @EricGustin suggests releasing in the
following order:
    1. `arcade-core` version 0.1.0
    2. `arcade-serve` version 0.1.0 and `arcade-tdk` version 0.1.0
    3. `arcade-ai` version 2.0.0
4. Patch release for all toolkits (all changes in toolkits are internal
refactors)
- [ ] [Update docs](https://github.com/ArcadeAI/docs/pull/318)

---------

Co-authored-by: Eric Gustin <eric@arcade.dev>
Co-authored-by: Eric Gustin <34000337+EricGustin@users.noreply.github.com>
2025-06-11 16:48:17 -07:00

240 lines
7 KiB
Python

from arcade_evals import (
EvalRubric,
EvalSuite,
ExpectedToolCall,
NumericCritic,
SimilarityCritic,
tool_eval,
)
from arcade_tdk import ToolCatalog
import arcade_search
from arcade_search.tools import search_google
# Evaluation rubric
rubric = EvalRubric(
fail_threshold=0.8,
warn_threshold=0.9,
)
catalog = ToolCatalog()
# Register the Google Search tool
catalog.add_module(arcade_search)
@tool_eval()
def google_search_eval_suite() -> EvalSuite:
"""Create an evaluation suite for the Google Search tool."""
suite = EvalSuite(
name="Google Search Tool Evaluation",
system_message="You are an AI assistant that can perform web searches using the provided tools.",
catalog=catalog,
rubric=rubric,
)
# Simple search query with default results
suite.add_case(
name="Simple search query with default results",
user_message="Search for 'Climate change effects on polar bears' on Google.",
expected_tool_calls=[
ExpectedToolCall(
func=search_google,
args={
"query": "Climate change effects on polar bears",
"n_results": 5,
},
)
],
critics=[
SimilarityCritic(critic_field="query", weight=1.0),
],
)
# Search query with specific number of results
suite.add_case(
name="Search query with specific number of results",
user_message="Find the top 3 articles about quantum computing.",
expected_tool_calls=[
ExpectedToolCall(
func=search_google,
args={
"query": "articles about quantum computing",
"n_results": 3,
},
)
],
critics=[
SimilarityCritic(critic_field="query", weight=0.7),
NumericCritic(
critic_field="n_results",
weight=0.3,
value_range=(1, 100),
),
],
)
# Search query with 'n' results specified in words
suite.add_case(
name="Search query with 'n' results specified in words",
user_message="Give me five recipes for vegan lasagna.",
expected_tool_calls=[
ExpectedToolCall(
func=search_google,
args={
"query": "recipes for vegan lasagna",
"n_results": 5,
},
)
],
critics=[
SimilarityCritic(critic_field="query", weight=0.7),
NumericCritic(
critic_field="n_results",
weight=0.3,
value_range=(1, 100),
),
],
)
# Ambiguous number of results
suite.add_case(
name="Ambiguous number of results",
user_message="Find articles about climate change impacts 10.",
expected_tool_calls=[
ExpectedToolCall(
func=search_google,
args={
"query": "articles about climate change impacts 10",
"n_results": 5,
},
)
],
critics=[
SimilarityCritic(critic_field="query", weight=1.0),
],
)
# Search query with multiple instructions
suite.add_case(
name="Search query with multiple instructions",
user_message="Search for the latest news on electric cars, and tell me about Tesla's new model.",
expected_tool_calls=[
ExpectedToolCall(
func=search_google,
args={
"query": "latest news on electric cars",
"n_results": 5,
},
),
ExpectedToolCall(
func=search_google,
args={
"query": "Tesla's new model",
"n_results": 5,
},
),
],
critics=[
SimilarityCritic(critic_field="query", weight=1.0),
],
)
# Search with stop words and filler words
suite.add_case(
name="Search with stop words and filler words",
user_message="Could you please search for the best ways to learn French?",
expected_tool_calls=[
ExpectedToolCall(
func=search_google,
args={
"query": "best ways to learn French",
"n_results": 5,
},
)
],
critics=[
SimilarityCritic(critic_field="query", weight=1.0),
],
)
# No clear query given
suite.add_case(
name="No clear query given",
user_message="Find it for me.",
expected_tool_calls=[],
critics=[],
)
# Search query with special characters
suite.add_case(
name="Search query with special characters",
user_message="Find me '@OpenAI's latest research papers'",
expected_tool_calls=[
ExpectedToolCall(
func=search_google,
args={
"query": "@OpenAI's latest research papers",
"n_results": 5,
},
)
],
critics=[
SimilarityCritic(critic_field="query", weight=1.0),
],
)
# Search query with complex instructions
suite.add_case(
name="Search query with complex instructions",
user_message="I need information about the impact of deforestation in the Amazon over the past decade.",
expected_tool_calls=[
ExpectedToolCall(
func=search_google,
args={
"query": "impact of deforestation in the Amazon over the past decade",
"n_results": 5,
},
)
],
critics=[
SimilarityCritic(critic_field="query", weight=1.0),
],
)
# Search query in a different language
suite.add_case(
name="Search query in a different language",
user_message="Busca información sobre la economía de España.",
expected_tool_calls=[
ExpectedToolCall(
func=search_google,
args={
"query": "economía de España",
"n_results": 5,
},
)
],
critics=[
SimilarityCritic(critic_field="query", weight=1.0),
],
)
# Search query with numeric data
suite.add_case(
name="Search query with numeric data",
user_message="What was the population of Japan in 2020?",
expected_tool_calls=[
ExpectedToolCall(
func=search_google,
args={
"query": "population of Japan in 2020",
"n_results": 5,
},
)
],
critics=[
SimilarityCritic(critic_field="query", weight=1.0),
],
)
return suite