ed89af4b4d
# Backwards-compatible refactoring of the Slack toolkit Several performance improvements, streamlined tool set, and easier to understand tool interfaces. All "old" tools were kept for backwards compatibility, with the same interfaces and response structure (but using the new and more performant tools under the hood). Full revision of unit-tests and evals. ## Streamlined tool set Multiple groups of tools were streamlined into a single one: * "get conversation metadata" from 5 tools to one; * "send message" from 2 tools to one; * "get users in conversation" from 3 tools to one; * "get messages" from 4 tools to one ## New capabilities * Messages retrieved are now populated with the users' names, apart from ID: makes it easier for LLMs to reference who sent a message, were mentioned, or reacted to a message * Retrieve users by username, email, and/or ID (before we only supported ID) * Retrieve multiple users in a single tool call ## Concurrency controls All operations issuing multiple requests concurrently now have a `Semaphore` to limit the concurrency level. The limit can be controlled through the `SLACK_MAX_CONCURRENT_REQUESTS` env var (defaults to 3). ## Networking performance improvement Various operations that used to make multiple API calls are now executed more efficiently: ### Find users by username * Before: a full scan of `users_list` was required (potentially multiple pages for large workspaces); * Now it stops as soon as we have all users needed (yes, it was dumb before) ### Get multiple users by their IDs * Before: for each user ID, we made one API call to the `users_info` endpoint * Now: we call `list_users` and scan the results to match the user IDs (an estimate of 99.5% of Slack workspaces have < 200 users; for large workspaces, we may need to paginate `list_users`) ### Get a conversation by its users * Before: * Call to `list_conversations` (potentially paginating) * For each conversation, one call to `conversations_members` (potentially paginating) * Then loop and find which conversation matches the users' IDs * Now: * A single call to `conversations_open`
259 lines
8.3 KiB
Python
259 lines
8.3 KiB
Python
import json
|
|
|
|
from arcade_evals import (
|
|
BinaryCritic,
|
|
EvalRubric,
|
|
EvalSuite,
|
|
ExpectedToolCall,
|
|
tool_eval,
|
|
)
|
|
from arcade_tdk import ToolCatalog
|
|
|
|
import arcade_slack
|
|
from arcade_slack.tools.users import (
|
|
get_users_info,
|
|
list_users,
|
|
)
|
|
|
|
# Evaluation rubric
|
|
rubric = EvalRubric(
|
|
fail_threshold=0.8,
|
|
warn_threshold=0.9,
|
|
)
|
|
|
|
|
|
catalog = ToolCatalog()
|
|
# Register the Slack tools
|
|
catalog.add_module(arcade_slack)
|
|
|
|
|
|
@tool_eval()
|
|
def get_user_info_eval_suite() -> EvalSuite:
|
|
"""Create an evaluation suite for tools getting user info by id."""
|
|
suite = EvalSuite(
|
|
name="Slack Users Tools Evaluation",
|
|
system_message="You are an AI assistant that can interact with Slack to get information about users.",
|
|
catalog=catalog,
|
|
rubric=rubric,
|
|
)
|
|
|
|
expected_user_id = "U12345"
|
|
|
|
get_user_info_by_id_eval_cases = [
|
|
(
|
|
"get user info by id",
|
|
f"What is the name of the user with id {expected_user_id}?",
|
|
),
|
|
(
|
|
"get user info by id",
|
|
f"get information about the user with id {expected_user_id}",
|
|
),
|
|
]
|
|
|
|
for name, user_message in get_user_info_by_id_eval_cases:
|
|
suite.add_case(
|
|
name=name,
|
|
user_message=user_message,
|
|
expected_tool_calls=[
|
|
ExpectedToolCall(
|
|
func=get_users_info,
|
|
args={
|
|
"user_ids": [expected_user_id],
|
|
"usernames": None,
|
|
"emails": None,
|
|
},
|
|
)
|
|
],
|
|
critics=[
|
|
BinaryCritic(critic_field="user_ids", weight=1 / 3),
|
|
BinaryCritic(critic_field="usernames", weight=1 / 3),
|
|
BinaryCritic(critic_field="emails", weight=1 / 3),
|
|
],
|
|
)
|
|
|
|
suite.add_case(
|
|
name="get user by username",
|
|
user_message="get the user 'johndoe'",
|
|
expected_tool_calls=[
|
|
ExpectedToolCall(
|
|
func=get_users_info,
|
|
args={
|
|
"usernames": ["johndoe"],
|
|
"user_ids": None,
|
|
"emails": None,
|
|
},
|
|
),
|
|
],
|
|
critics=[
|
|
BinaryCritic(critic_field="usernames", weight=1 / 3),
|
|
BinaryCritic(critic_field="user_ids", weight=1 / 3),
|
|
BinaryCritic(critic_field="emails", weight=1 / 3),
|
|
],
|
|
)
|
|
|
|
suite.add_case(
|
|
name="get user by email",
|
|
user_message="get the user 'john.doe@acme.com'",
|
|
expected_tool_calls=[
|
|
ExpectedToolCall(
|
|
func=get_users_info,
|
|
args={
|
|
"usernames": None,
|
|
"user_ids": None,
|
|
"emails": ["john.doe@acme.com"],
|
|
},
|
|
),
|
|
],
|
|
critics=[
|
|
BinaryCritic(critic_field="usernames", weight=1 / 3),
|
|
BinaryCritic(critic_field="user_ids", weight=1 / 3),
|
|
BinaryCritic(critic_field="emails", weight=1 / 3),
|
|
],
|
|
)
|
|
|
|
suite.add_case(
|
|
name="get multiple users by username",
|
|
user_message="get the users with the usernames 'johndoe' and 'foobar'",
|
|
expected_tool_calls=[
|
|
ExpectedToolCall(
|
|
func=get_users_info,
|
|
args={
|
|
"usernames": ["johndoe", "foobar"],
|
|
"user_ids": None,
|
|
"emails": None,
|
|
},
|
|
),
|
|
],
|
|
critics=[
|
|
BinaryCritic(critic_field="usernames", weight=1 / 3),
|
|
BinaryCritic(critic_field="user_ids", weight=1 / 3),
|
|
BinaryCritic(critic_field="emails", weight=1 / 3),
|
|
],
|
|
)
|
|
|
|
suite.add_case(
|
|
name="get multiple users by email",
|
|
user_message="get the users with the emails 'john.doe@acme.com' and 'jane.doe@acme.com'",
|
|
expected_tool_calls=[
|
|
ExpectedToolCall(
|
|
func=get_users_info,
|
|
args={
|
|
"usernames": None,
|
|
"user_ids": None,
|
|
"emails": ["john.doe@acme.com", "jane.doe@acme.com"],
|
|
},
|
|
),
|
|
],
|
|
critics=[
|
|
BinaryCritic(critic_field="usernames", weight=1 / 3),
|
|
BinaryCritic(critic_field="user_ids", weight=1 / 3),
|
|
BinaryCritic(critic_field="emails", weight=1 / 3),
|
|
],
|
|
)
|
|
|
|
return suite
|
|
|
|
|
|
@tool_eval()
|
|
def list_users_eval_suite() -> EvalSuite:
|
|
"""Create an evaluation suite for tools listing users."""
|
|
suite = EvalSuite(
|
|
name="Slack Users Tools Evaluation",
|
|
system_message="You are an AI assistant that can interact with Slack to get information about users.",
|
|
catalog=catalog,
|
|
rubric=rubric,
|
|
)
|
|
|
|
suite.add_case(
|
|
name="list users",
|
|
user_message="list all users on my slack workspace",
|
|
expected_tool_calls=[
|
|
ExpectedToolCall(func=list_users, args={}),
|
|
],
|
|
)
|
|
|
|
suite.add_case(
|
|
name="list users without bots",
|
|
user_message="list all users on my slack workspace, except bots",
|
|
expected_tool_calls=[
|
|
ExpectedToolCall(func=list_users, args={"exclude_bots": True}),
|
|
],
|
|
critics=[
|
|
BinaryCritic(critic_field="exclude_bots", weight=1.0),
|
|
],
|
|
)
|
|
|
|
suite.add_case(
|
|
name="list 10 users without bots",
|
|
user_message="get a list of 10 users on my slack workspace, except bots",
|
|
expected_tool_calls=[
|
|
ExpectedToolCall(func=list_users, args={"exclude_bots": True, "limit": 10}),
|
|
],
|
|
critics=[
|
|
BinaryCritic(critic_field="exclude_bots", weight=0.5),
|
|
BinaryCritic(critic_field="limit", weight=0.5),
|
|
],
|
|
)
|
|
|
|
suite.add_case(
|
|
name="test list users with pagination",
|
|
user_message="get the next 5 users",
|
|
expected_tool_calls=[
|
|
ExpectedToolCall(
|
|
func=list_users,
|
|
args={"limit": 5, "next_cursor": "dXNlcjpVsDjzOTZGVDlQRA=="},
|
|
),
|
|
],
|
|
critics=[
|
|
BinaryCritic(critic_field="limit", weight=0.5),
|
|
BinaryCritic(critic_field="next_cursor", weight=0.5),
|
|
],
|
|
additional_messages=[
|
|
{"role": "user", "content": "get a list of 2 users from my slack workspace"},
|
|
{
|
|
"role": "assistant",
|
|
"content": "",
|
|
"tool_calls": [
|
|
{
|
|
"id": "call_1",
|
|
"type": "function",
|
|
"function": {"name": "Slack_ListUsers", "arguments": '{"limit":2}'},
|
|
}
|
|
],
|
|
},
|
|
{
|
|
"role": "tool",
|
|
"content": json.dumps({
|
|
"next_cursor": "dXNlcjpVsDjzOTZGVDlQRA==",
|
|
"users": [
|
|
{
|
|
"display_name": "John Doe",
|
|
"email": "john.doe@acme.com",
|
|
"id": "U123",
|
|
"is_bot": False,
|
|
"name": "john.doe",
|
|
"real_name": "John Doe",
|
|
"timezone": "America/Los_Angeles",
|
|
},
|
|
{
|
|
"display_name": "Jane Doe",
|
|
"email": "jane.doe@acme.com",
|
|
"id": "U124",
|
|
"is_bot": False,
|
|
"name": "jane.doe",
|
|
"real_name": "Jane Doe",
|
|
"timezone": "America/Los_Angeles",
|
|
},
|
|
],
|
|
}),
|
|
"tool_call_id": "call_1",
|
|
"name": "Slack_ListUsers",
|
|
},
|
|
{
|
|
"role": "assistant",
|
|
"content": "Here are two users from your Slack workspace:\n\n1. **John Doe**\n - Display Name: John Doe\n - Email: john.doe@acme.com\n - Timezone: America/Los_Angeles\n\n2. **Jane Doe**\n - Display Name: Jane Doe\n - Email: jane.doe@acme.com\n - Timezone: America/Los_Angeles\n\nIf you need more information or additional users, feel free to ask!",
|
|
},
|
|
],
|
|
)
|
|
|
|
return suite
|