arcade-mcp/toolkits/spotify/evals/eval_player.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

174 lines
5.3 KiB
Python

from arcade_evals import (
BinaryCritic,
EvalRubric,
EvalSuite,
ExpectedToolCall,
NumericCritic,
SimilarityCritic,
tool_eval,
)
from arcade_tdk import ToolCatalog
from arcade_spotify.tools.player import (
adjust_playback_position,
get_currently_playing,
get_playback_state,
pause_playback,
play_artist_by_name,
play_track_by_name,
resume_playback,
skip_to_next_track,
skip_to_previous_track,
start_tracks_playback_by_id,
)
# Evaluation rubric
rubric = EvalRubric(
fail_threshold=0.9,
warn_threshold=0.95,
)
catalog = ToolCatalog()
catalog.add_tool(adjust_playback_position, "Spotify")
catalog.add_tool(skip_to_next_track, "Spotify")
catalog.add_tool(skip_to_previous_track, "Spotify")
catalog.add_tool(pause_playback, "Spotify")
catalog.add_tool(resume_playback, "Spotify")
catalog.add_tool(start_tracks_playback_by_id, "Spotify")
catalog.add_tool(get_playback_state, "Spotify")
catalog.add_tool(get_currently_playing, "Spotify")
catalog.add_tool(play_artist_by_name, "Spotify")
catalog.add_tool(play_track_by_name, "Spotify")
@tool_eval()
def spotify_player_eval_suite() -> EvalSuite:
"""Create an evaluation suite for Spotify "player" tools."""
suite = EvalSuite(
name="Spotify Tools Evaluation",
system_message="You are an AI assistant that can manage Spotify using the provided tools.",
catalog=catalog,
rubric=rubric,
)
suite.add_case(
name="Adjust playback position",
user_message="can you skip to the 10th second of the song",
expected_tool_calls=[
ExpectedToolCall(
func=adjust_playback_position,
args={"absolute_position_ms": 10000},
)
],
critics=[
NumericCritic(
critic_field="absolute_position_ms", weight=1.0, value_range=(9000, 11000)
),
],
)
suite.add_case(
name="Adjust playback position relative to current position",
user_message="go back 10 seconds",
expected_tool_calls=[
ExpectedToolCall(
func=adjust_playback_position,
args={"relative_position_ms": -10000},
)
],
critics=[
NumericCritic(
critic_field="relative_position_ms",
weight=1.0,
value_range=(-11000, -9000),
),
],
)
suite.add_case(
name="Skip to previous track",
user_message="oops i didn't mean to skip that song, go back",
expected_tool_calls=[ExpectedToolCall(func=skip_to_previous_track, args={})],
)
suite.add_case(
name="Skip to next track",
user_message="skip this song and also the next one",
expected_tool_calls=[
ExpectedToolCall(func=skip_to_next_track, args={}),
ExpectedToolCall(func=skip_to_next_track, args={}),
],
)
suite.add_case(
name="Pause playback",
user_message="wait im getting a text, stop playing it please",
expected_tool_calls=[ExpectedToolCall(func=pause_playback, args={})],
)
suite.add_case(
name="Resume playback",
user_message="ok i'm back, you can press play again",
expected_tool_calls=[ExpectedToolCall(func=resume_playback, args={})],
)
suite.add_case(
name="Start playback of a list of tracks",
user_message="Play these two 03gaqN3aWm9TQxuHay0G8R, 03gaqN3aWm9TQxuHay0G8R. But start at the 10th second of the first track",
expected_tool_calls=[
ExpectedToolCall(
func=start_tracks_playback_by_id,
args={
"track_ids": ["03gaqN3aWm9TQxuHay0G8R", "03gaqN3aWm9TQxuHay0G8R"],
"position_ms": 10000,
},
)
],
critics=[
BinaryCritic(critic_field="track_ids", weight=0.5),
NumericCritic(critic_field="position_ms", weight=0.5, value_range=(9000, 11000)),
],
)
suite.add_case(
name="Get playback state",
user_message="what's the name of this song and who plays it?",
expected_tool_calls=[ExpectedToolCall(func=get_currently_playing, args={})],
)
suite.add_case(
name="Get playback state",
user_message="what device is playing music rn?",
expected_tool_calls=[ExpectedToolCall(func=get_playback_state, args={})],
)
suite.add_case(
name="Play artist by name",
user_message="play pearl jam",
expected_tool_calls=[
ExpectedToolCall(
func=play_artist_by_name,
args={"name": "Pearl Jam"},
)
],
critics=[
SimilarityCritic(critic_field="name", weight=1.0),
],
)
suite.add_case(
name="Play track by name",
user_message="it would be really great if I could listen to strobe by deadmau5 right now.",
expected_tool_calls=[
ExpectedToolCall(
func=play_track_by_name,
args={"track_name": "strobe", "artist_name": "deadmau5"},
)
],
critics=[
SimilarityCritic(critic_field="track_name", weight=0.5),
SimilarityCritic(critic_field="artist_name", weight=0.5),
],
)
return suite