Merge branch 'main' into fix_graph

This commit is contained in:
seanzhou1023
2025-07-11 12:55:24 -07:00
committed by GitHub
85 changed files with 8155 additions and 618 deletions
+69
View File
@@ -1,5 +1,74 @@
# Changelog
## [1.6.1](https://github.com/google/adk-python/compare/v1.5.0...v1.6.1) (2025-07-09)
### Features
* Add A2A support as experimental features [f0183a9](https://github.com/google/adk-python/commit/f0183a9b98b0bcf8aab4f948f467cef204ddc9d6)
* Install google-adk with a2a extra: pip install google-adk[a2a]
* Users can serve agents as A2A agent with `--a2a` option for `adk web` and
`adk api_server`
* Users can run a remote A2A agent with `RemoteA2AAgent` class
* Three A2A agent samples are added:
* contributing/samples/a2a_basic
* contributing/samples/a2a_auth
* contributing/samples/a2a_human_in_loop
* Support agent hot reload.[e545e5a](https://github.com/google/adk-python/commit/e545e5a570c1331d2ed8fda31c7244b5e0f71584)
Users can add `--reload_agents` flag to `adk web` and `adk api_server` command
to reload agents automatically when new changes are detected.
* Eval features
* Implement auto rater-based evaluator for responses [75699fb](https://github.com/google/adk-python/commit/75699fbeca06f99c6f2415938da73bb423ec9b9b)
* Add Safety evaluator metric [0bd05df](https://github.com/google/adk-python/commit/0bd05df471a440159a44b5864be4740b0f1565f9)
* Add BaseEvalService declaration and surrounding data models [b0d88bf](https://github.com/google/adk-python/commit/b0d88bf17242e738bcd409b3d106deed8ce4d407)
* Minor features
* Add `custom_metadata` to VertexAiSessionService when adding events [a021222](https://github.com/google/adk-python/commit/a02122207734cabb26f7c23e84d2336c4b8b0375)
* Support protected write in BigQuery `execute_sql` tool [dc43d51](https://github.com/google/adk-python/commit/dc43d518c90b44932b3fdedd33fca9e6c87704e2)
* Added clone() method to BaseAgent to allow users to create copies of an agent [d263afd] (https://github.com/google/adk-python/commit/d263afd91ba4a3444e5321c0e1801c499dec4c68)
### Bug Fixes
* Support project-based gemini model path to use enterprise_web_search_tool [e33161b](https://github.com/google/adk-python/commit/e33161b4f8650e8bcb36c650c4e2d1fe79ae2526)
* Use inspect.signature() instead of typing.get_type_hints for examining function signatures[4ca77bc](https://github.com/google/adk-python/commit/4ca77bc056daa575621a80d3c8d5014b78209233)
* Replace Event ID generation with UUID4 to prevent SQLite integrity constraint failures [e437c7a](https://github.com/google/adk-python/commit/e437c7aac650ac6a53fcfa71bd740e3e5ec0f230)
* Remove duplicate options from `adk deploy` [3fa2ea7](https://github.com/google/adk-python/commit/3fa2ea7cb923c9f8606d98b45a23bd58a7027436)
* Fix scenario where a user can access another users events given the same session id [362fb3f](https://github.com/google/adk-python/commit/362fb3f2b7ac4ad15852d00ce4f3935249d097f6)
* Handle unexpected 'parameters' argument in FunctionTool.run_async [0959b06](https://github.com/google/adk-python/commit/0959b06dbdf3037fe4121f12b6d25edca8fb9afc)
* Make sure each partial event has different timestamp [17d6042](https://github.com/google/adk-python/commit/17d604299505c448fcb55268f0cbaeb6c4fa314a)
* Avoid pydantic.ValidationError when the model stream returns empty final chunk [9b75e24](https://github.com/google/adk-python/commit/9b75e24d8c01878c153fec26ccfea4490417d23b)
* Fix google_search_tool.py to support updated Gemini LIVE model naming [77b869f](https://github.com/google/adk-python/commit/77b869f5e35a66682cba35563824fd23a9028d7c)
* Adding detailed information on each metric evaluation [04de3e1](https://github.com/google/adk-python/commit/04de3e197d7a57935488eb7bfa647c7ab62cd9d9)
* Converts litellm generate config err [3901fad](https://github.com/google/adk-python/commit/3901fade71486a1e9677fe74a120c3f08efe9d9e)
* Save output in state via output_key only when the event is authored by current agent [20279d9](https://github.com/google/adk-python/commit/20279d9a50ac051359d791dea77865c17c0bbf9e)
* Treat SQLite database update time as UTC for session's last update time [3f621ae](https://github.com/google/adk-python/commit/3f621ae6f2a5fac7f992d3d833a5311b4d4e7091)
* Raise ValueError when sessionId and userId are incorrect combination(#1653) [4e765ae](https://github.com/google/adk-python/commit/4e765ae2f3821318e581c26a52e11d392aaf72a4)
* Support API-Key for MCP Tool authentication [045aea9](https://github.com/google/adk-python/commit/045aea9b15ad0190a960f064d6e1e1fc7f964c69)
* Lock LangGraph version to <= 0.4.10 [9029b8a](https://github.com/google/adk-python/commit/9029b8a66e9d5e0d29d9a6df0e5590cc7c0e9038)
* Update the retry logic of create session polling [3d2f13c](https://github.com/google/adk-python/commit/3d2f13cecd3fef5adfa1c98bf23d7b68ff355f4d)
### Chores
* Extract mcp client creation logic to a separate method [45d60a1](https://github.com/google/adk-python/commit/45d60a1906bfe7c43df376a829377e2112ea3d17)
* Add tests for live streaming configs [bf39c00](https://github.com/google/adk-python/commit/bf39c006102ef3f01e762e7bb744596a4589f171)
* Update ResponseEvaluator to use newer version of Eval SDK [62c4a85](https://github.com/google/adk-python/commit/62c4a8591780a9a3fdb03a0de11092d84118a1b9)
* Add util to build our llms.txt and llms-full.txt files [a903c54](https://github.com/google/adk-python/commit/a903c54bacfcb150dc315bec9c67bf7ce9551c07)
* Create an example for multi agent live streaming [a58cc3d](https://github.com/google/adk-python/commit/a58cc3d882e59358553e8ea16d166b1ab6d3aa71)
* Refactor the ADK Triaging Agent to make the code easier to read [b6c7b5b](https://github.com/google/adk-python/commit/b6c7b5b64fcd2e83ed43f7b96ea43791733955d8)
### Documentation
* Update the a2a exmaple link in README.md [d0fdfb8](https://github.com/google/adk-python/commit/d0fdfb8c8e2e32801999c81de8d8ed0be3f88e76)
* Adds AGENTS.md to provide relevant project context for the Gemini CLI [37108be](https://github.com/google/adk-python/commit/37108be8557e011f321de76683835448213f8515)
* Update CONTRIBUTING.md [ffa9b36](https://github.com/google/adk-python/commit/ffa9b361db615ae365ba62c09a8f4226fb761551)
* Add adk project overview and architecture [28d0ea8](https://github.com/google/adk-python/commit/28d0ea876f2f8de952f1eccbc788e98e39f50cf5)
* Add docstring to clarify that inmemory service are not suitable for production [dc414cb](https://github.com/google/adk-python/commit/dc414cb5078326b8c582b3b9072cbda748766286)
* Update agents.md to include versioning strategy [6a39c85](https://github.com/google/adk-python/commit/6a39c854e032bda3bc15f0e4fe159b41cf2f474b)
* Add tenacity into project.toml [df141db](https://github.com/google/adk-python/commit/df141db60c1137a6bcddd6d46aad3dc506868543)
* Updating CONTRIBUTING.md with missing extra [e153d07](https://github.com/google/adk-python/commit/e153d075939fb628a7dc42b12e1b3461842db541)
## [1.5.0](https://github.com/google/adk-python/compare/v1.4.2...v1.5.0) (2025-06-25)
+1 -1
View File
@@ -93,7 +93,7 @@ prime_agent = RemoteA2aAgent(
root_agent = Agent(
model="gemini-1.5-flash",
model="gemini-2.0-flash",
name="root_agent",
instruction="""
You are a helpful assistant that can roll dice and check if numbers are prime.
@@ -33,7 +33,7 @@ approval_agent = RemoteA2aAgent(
root_agent = Agent(
model='gemini-1.5-flash',
model='gemini-2.0-flash',
name='reimbursement_agent',
instruction="""
You are an agent whose job is to handle the reimbursement process for
@@ -39,7 +39,7 @@ def ask_for_approval(
root_agent = Agent(
model='gemini-1.5-flash',
model='gemini-2.0-flash',
name='reimbursement_agent',
instruction="""
You are an agent whose job is to handle the reimbursement process for
@@ -0,0 +1,83 @@
# ADK Answering Agent
The ADK Answering Agent is a Python-based agent designed to help answer questions in GitHub discussions for the `google/adk-python` repository. It uses a large language model to analyze open discussions, retrieve information from document store, generate response, and post a comment in the github discussion.
This agent can be operated in three distinct modes: an interactive mode for local use, a batch script mode for oncall use, or as a fully automated GitHub Actions workflow (TBD).
---
## Interactive Mode
This mode allows you to run the agent locally to review its recommendations in real-time before any changes are made to your repository's issues.
### Features
* **Web Interface**: The agent's interactive mode can be rendered in a web browser using the ADK's `adk web` command.
* **User Approval**: In interactive mode, the agent is instructed to ask for your confirmation before posting a comment to a GitHub issue.
* **Question & Answer**: You can ask ADK related questions, and the agent will provide answers based on its knowledge on ADK.
### Running in Interactive Mode
To run the agent in interactive mode, first set the required environment variables. Then, execute the following command in your terminal:
```bash
adk web
```
This will start a local server and provide a URL to access the agent's web interface in your browser.
---
## Batch Script Mode
The `answer_discussions.py` is created for ADK oncall team to batch process discussions.
### Features
* **Batch Process**: Taken either a number as the count of the recent discussions or a list of discussion numbers, the script will invoke the agent to answer all the specified discussions in one single run.
### Running in Interactive Mode
To run the agent in batch script mode, first set the required environment variables. Then, execute the following command in your terminal:
```bash
export PYTHONPATH=contributing/samples
python -m adk_answering_agent.answer_discussions --numbers 27 36 # Answer specified discussions
```
Or `python -m adk_answering_agent.answer_discussions --recent 10` to answer the 10 most recent updated discussions.
---
## GitHub Workflow Mode
The `main.py` is reserved for the Github Workflow. The detailed setup for the automatic workflow is TBD.
---
## Setup and Configuration
Whether running in interactive or workflow mode, the agent requires the following setup.
### Dependencies
The agent requires the following Python libraries.
```bash
pip install --upgrade pip
pip install google-adk requests
```
The agent also requires gcloud login:
```bash
gcloud auth application-default login
```
### Environment Variables
The following environment variables are required for the agent to connect to the necessary services.
* `GITHUB_TOKEN=YOUR_GITHUB_TOKEN`: **(Required)** A GitHub Personal Access Token with `issues:write` permissions. Needed for both interactive and workflow modes.
* `GOOGLE_GENAI_USE_VERTEXAI=TRUE`: **(Required)** Use Google Vertex AI for the authentication.
* `GOOGLE_CLOUD_PROJECT=YOUR_PROJECT_ID`: **(Required)** The Google Cloud project ID.
* `GOOGLE_CLOUD_LOCATION=LOCATION`: **(Required)** The Google Cloud region.
* `VERTEXAI_DATASTORE_ID=YOUR_DATASTORE_ID`: **(Required)** The Vertex AI datastore ID for the document store (i.e. knowledge base).
* `OWNER`: The GitHub organization or username that owns the repository (e.g., `google`). Needed for both modes.
* `REPO`: The name of the GitHub repository (e.g., `adk-python`). Needed for both modes.
* `INTERACTIVE`: Controls the agent's interaction mode. For the automated workflow, this is set to `0`. For interactive mode, it should be set to `1` or left unset.
For local execution in interactive mode, you can place these variables in a `.env` file in the project's root directory. For the GitHub workflow, they should be configured as repository secrets.
@@ -0,0 +1,15 @@
# Copyright 2025 Google LLC
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
from . import agent
@@ -0,0 +1,192 @@
# Copyright 2025 Google LLC
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
from typing import Any
from adk_answering_agent.settings import IS_INTERACTIVE
from adk_answering_agent.settings import OWNER
from adk_answering_agent.settings import REPO
from adk_answering_agent.settings import VERTEXAI_DATASTORE_ID
from adk_answering_agent.utils import error_response
from adk_answering_agent.utils import run_graphql_query
from google.adk.agents import Agent
from google.adk.tools import VertexAiSearchTool
import requests
if IS_INTERACTIVE:
APPROVAL_INSTRUCTION = (
"Ask for user approval or confirmation for adding the comment."
)
else:
APPROVAL_INSTRUCTION = (
"**Do not** wait or ask for user approval or confirmation for adding the"
" comment."
)
def get_discussion_and_comments(discussion_number: int) -> dict[str, Any]:
"""Fetches a discussion and its comments using the GitHub GraphQL API.
Args:
discussion_number: The number of the GitHub discussion.
Returns:
A dictionary with the request status and the discussion details.
"""
print(f"Attempting to get discussion #{discussion_number} and its comments")
query = """
query($owner: String!, $repo: String!, $discussionNumber: Int!) {
repository(owner: $owner, name: $repo) {
discussion(number: $discussionNumber) {
id
title
body
createdAt
closed
author {
login
}
# For each comment, fetch the latest 100 comments.
comments(last: 100) {
nodes {
id
body
createdAt
author {
login
}
# For each comment, fetch the latest 50 replies
replies(last: 50) {
nodes {
id
body
createdAt
author {
login
}
}
}
}
}
}
}
}
"""
variables = {
"owner": OWNER,
"repo": REPO,
"discussionNumber": discussion_number,
}
try:
response = run_graphql_query(query, variables)
if "errors" in response:
return error_response(str(response["errors"]))
discussion_data = (
response.get("data", {}).get("repository", {}).get("discussion")
)
if not discussion_data:
return error_response(f"Discussion #{discussion_number} not found.")
return {"status": "success", "discussion": discussion_data}
except requests.exceptions.RequestException as e:
return error_response(str(e))
def add_comment_to_discussion(
discussion_id: str, comment_body: str
) -> dict[str, Any]:
"""Adds a comment to a specific discussion.
Args:
discussion_id: The GraphQL node ID of the discussion.
comment_body: The content of the comment in Markdown.
Returns:
The status of the request and the new comment's details.
"""
print(f"Adding comment to discussion {discussion_id}")
query = """
mutation($discussionId: ID!, $body: String!) {
addDiscussionComment(input: {discussionId: $discussionId, body: $body}) {
comment {
id
body
createdAt
author {
login
}
}
}
}
"""
variables = {"discussionId": discussion_id, "body": comment_body}
try:
response = run_graphql_query(query, variables)
if "errors" in response:
return error_response(str(response["errors"]))
new_comment = (
response.get("data", {}).get("addDiscussionComment", {}).get("comment")
)
return {"status": "success", "comment": new_comment}
except requests.exceptions.RequestException as e:
return error_response(str(e))
root_agent = Agent(
model="gemini-2.5-pro",
name="adk_answering_agent",
description="Answer questions about ADK repo.",
instruction=f"""
You are a helpful assistant that responds to questions from the GitHub repository `{OWNER}/{REPO}`
based on information about Google ADK found in the document store. You can access the document store
using the `VertexAiSearchTool`.
When user specifies a discussion number, here are the steps:
1. Use the `get_discussion_and_comments` tool to get the details of the discussion including the comments.
2. Focus on the latest comment but reference all comments if needed to understand the context.
* If there is no comment at all, just focus on the discussion title and body.
3. If all the following conditions are met, try to add a comment to the discussion, otherwise, do not respond:
* The discussion is not closed.
* The latest comment is not from you or other agents (marked as "Response from XXX Agent").
* The latest comment is asking a question or requesting information.
4. Use the `VertexAiSearchTool` to find relevant information before answering.
IMPORTANT:
* {APPROVAL_INSTRUCTION}
* Your response should be based on the information you found in the document store. Do not invent
information that is not in the document store. Do not invent citations which are not in the document store.
* If you can't find the answer or information in the document store, **do not** respond.
* Include a bolded note (e.g. "Response from ADK Answering Agent") in your comment
to indicate this comment was added by an ADK Answering Agent.
* Have an empty line between the note and the rest of your response.
* Inlclude a short summary of your response in the comment as a TLDR, e.g. "**TLDR**: <your summary>".
* Have a divider line between the TLDR and your detail response.
* Do not respond to any other discussion except the one specified by the user.
* Please include your justification for your decision in your output
to the user who is telling with you.
* If you uses citation from the document store, please provide a footnote
referencing the source document format it as: "[1] URL of the document".
* Replace the "gs://prefix/" part, e.g. "gs://adk-qa-bucket/", to be "https://github.com/google/"
* Add "blob/main/" after the repo name, e.g. "adk-python", "adk-docs", for example:
* If the original URL is "gs://adk-qa-bucket/adk-python/src/google/adk/version.py",
then the citation URL is "https://github.com/google/adk-python/blob/main/src/google/adk/version.py",
* If the original URL is "gs://adk-qa-bucket/adk-docs/docs/index.md",
then the citation URL is "https://github.com/google/adk-docs/blob/main/docs/index.md"
* If the file is a html file, replace the ".html" to be ".md"
""",
tools=[
VertexAiSearchTool(data_store_id=VERTEXAI_DATASTORE_ID),
get_discussion_and_comments,
add_comment_to_discussion,
],
)
@@ -0,0 +1,172 @@
# Copyright 2025 Google LLC
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
import argparse
import asyncio
import sys
import time
from adk_answering_agent import agent
from adk_answering_agent.settings import OWNER
from adk_answering_agent.settings import REPO
from adk_answering_agent.utils import call_agent_async
from adk_answering_agent.utils import run_graphql_query
from google.adk.runners import InMemoryRunner
import requests
APP_NAME = "adk_discussion_answering_app"
USER_ID = "adk_discussion_answering_assistant"
async def list_most_recent_discussions(count: int = 1) -> list[int] | None:
"""Fetches a specified number of the most recently updated discussions.
Args:
count: The number of discussions to retrieve. Defaults to 1.
Returns:
A list of discussion numbers.
"""
print(
f"Attempting to fetch the {count} most recently updated discussions from"
f" {OWNER}/{REPO}..."
)
query = """
query($owner: String!, $repo: String!, $count: Int!) {
repository(owner: $owner, name: $repo) {
discussions(
first: $count
orderBy: {field: UPDATED_AT, direction: DESC}
) {
nodes {
title
number
updatedAt
author {
login
}
}
}
}
}
"""
variables = {"owner": OWNER, "repo": REPO, "count": count}
try:
response = run_graphql_query(query, variables)
if "errors" in response:
print(f"Error from GitHub API: {response['errors']}", file=sys.stderr)
return None
discussions = (
response.get("data", {})
.get("repository", {})
.get("discussions", {})
.get("nodes", [])
)
return [d["number"] for d in discussions]
except requests.exceptions.RequestException as e:
print(f"Request failed: {e}", file=sys.stderr)
return None
def process_arguments():
"""Parses command-line arguments."""
parser = argparse.ArgumentParser(
description="A script that answer questions for Github discussions.",
epilog=(
"Example usage: \n"
"\tpython -m adk_answering_agent.answer_discussions --recent 10\n"
"\tpython -m adk_answering_agent.answer_discussions --numbers 21 31\n"
),
formatter_class=argparse.RawTextHelpFormatter,
)
group = parser.add_mutually_exclusive_group(required=True)
group.add_argument(
"--recent",
type=int,
metavar="COUNT",
help="Answer the N most recently updated discussion numbers.",
)
group.add_argument(
"--numbers",
type=int,
nargs="+",
metavar="NUM",
help="Answer a specific list of discussion numbers.",
)
if len(sys.argv) == 1:
parser.print_help(sys.stderr)
sys.exit(1)
return parser.parse_args()
async def main():
args = process_arguments()
discussion_numbers = []
if args.recent:
discussion_numbers = await list_most_recent_discussions(count=args.recent)
elif args.numbers:
discussion_numbers = args.numbers
if not discussion_numbers:
print("No discussions specified. Exiting...", file=sys.stderr)
sys.exit(1)
print(f"Will try to answer discussions: {discussion_numbers}...")
runner = InMemoryRunner(
agent=agent.root_agent,
app_name=APP_NAME,
)
for discussion_number in discussion_numbers:
print("#" * 80)
print(f"Starting to process discussion #{discussion_number}...")
# Create a new session for each discussion to avoid interference.
session = await runner.session_service.create_session(
app_name=APP_NAME, user_id=USER_ID
)
prompt = (
f"Please check discussion #{discussion_number} see if you can help"
" answer the question or provide some information!"
)
response = await call_agent_async(runner, USER_ID, session.id, prompt)
print(f"<<<< Agent Final Output: {response}\n")
if __name__ == "__main__":
start_time = time.time()
print(
f"Start answering discussions for {OWNER}/{REPO} at"
f" {time.strftime('%Y-%m-%d %H:%M:%S', time.gmtime(start_time))}"
)
print("-" * 80)
asyncio.run(main())
print("-" * 80)
end_time = time.time()
print(
"Discussion answering finished at"
f" {time.strftime('%Y-%m-%d %H:%M:%S', time.gmtime(end_time))}",
)
print("Total script execution time:", f"{end_time - start_time:.2f} seconds")
@@ -0,0 +1,66 @@
# Copyright 2025 Google LLC
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
import asyncio
import time
from adk_answering_agent import agent
from adk_answering_agent.settings import DISCUSSION_NUMBER
from adk_answering_agent.settings import OWNER
from adk_answering_agent.settings import REPO
from adk_answering_agent.utils import call_agent_async
from adk_answering_agent.utils import parse_number_string
from google.adk.runners import InMemoryRunner
APP_NAME = "adk_answering_app"
USER_ID = "adk_answering_user"
async def main():
runner = InMemoryRunner(
agent=agent.root_agent,
app_name=APP_NAME,
)
session = await runner.session_service.create_session(
app_name=APP_NAME, user_id=USER_ID
)
discussion_number = parse_number_string(DISCUSSION_NUMBER)
if not discussion_number:
print(f"Error: Invalid discussion number received: {DISCUSSION_NUMBER}.")
return
prompt = (
f"Please check discussion #{discussion_number} see if you can help answer"
" the question or provide some information!"
)
response = await call_agent_async(runner, USER_ID, session.id, prompt)
print(f"<<<< Agent Final Output: {response}\n")
if __name__ == "__main__":
start_time = time.time()
print(
f"Start Q&A checking on {OWNER}/{REPO} discussion #{DISCUSSION_NUMBER} at"
f" {time.strftime('%Y-%m-%d %H:%M:%S', time.gmtime(start_time))}"
)
print("-" * 80)
asyncio.run(main())
print("-" * 80)
end_time = time.time()
print(
"Q&A checking finished at"
f" {time.strftime('%Y-%m-%d %H:%M:%S', time.gmtime(end_time))}",
)
print("Total script execution time:", f"{end_time - start_time:.2f} seconds")
@@ -0,0 +1,36 @@
# Copyright 2025 Google LLC
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
import os
from dotenv import load_dotenv
load_dotenv(override=True)
GITHUB_BASE_URL = "https://api.github.com"
GITHUB_GRAPHQL_URL = GITHUB_BASE_URL + "/graphql"
GITHUB_TOKEN = os.getenv("GITHUB_TOKEN")
if not GITHUB_TOKEN:
raise ValueError("GITHUB_TOKEN environment variable not set")
VERTEXAI_DATASTORE_ID = os.getenv("VERTEXAI_DATASTORE_ID")
if not VERTEXAI_DATASTORE_ID:
raise ValueError("VERTEXAI_DATASTORE_ID environment variable not set")
OWNER = os.getenv("OWNER", "google")
REPO = os.getenv("REPO", "adk-python")
DISCUSSION_NUMBER = os.getenv("DISCUSSION_NUMBER")
IS_INTERACTIVE = os.getenv("INTERACTIVE", "1").lower() in ["true", "1"]
@@ -0,0 +1,81 @@
# Copyright 2025 Google LLC
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
import sys
from typing import Any
from adk_answering_agent.settings import GITHUB_GRAPHQL_URL
from adk_answering_agent.settings import GITHUB_TOKEN
from google.adk.agents.run_config import RunConfig
from google.adk.runners import Runner
from google.genai import types
import requests
headers = {
"Authorization": f"token {GITHUB_TOKEN}",
"Accept": "application/vnd.github.v3+json",
}
def error_response(error_message: str) -> dict[str, Any]:
return {"status": "error", "error_message": error_message}
def run_graphql_query(query: str, variables: dict[str, Any]) -> dict[str, Any]:
"""Executes a GraphQL query."""
payload = {"query": query, "variables": variables}
response = requests.post(
GITHUB_GRAPHQL_URL, headers=headers, json=payload, timeout=60
)
response.raise_for_status()
return response.json()
def parse_number_string(number_str: str | None, default_value: int = 0) -> int:
"""Parse a number from the given string."""
if not number_str:
return default_value
try:
return int(number_str)
except ValueError:
print(
f"Warning: Invalid number string: {number_str}. Defaulting to"
f" {default_value}.",
file=sys.stderr,
)
return default_value
async def call_agent_async(
runner: Runner, user_id: str, session_id: str, prompt: str
) -> str:
"""Call the agent asynchronously with the user's prompt."""
content = types.Content(
role="user", parts=[types.Part.from_text(text=prompt)]
)
final_response_text = ""
async for event in runner.run_async(
user_id=user_id,
session_id=session_id,
new_message=content,
run_config=RunConfig(save_input_blobs_as_artifacts=False),
):
if event.content and event.content.parts:
if text := "".join(part.text or "" for part in event.content.parts):
if event.author != "user":
final_response_text += text
return final_response_text
@@ -20,6 +20,7 @@ import requests
headers = {
"Authorization": f"token {GITHUB_TOKEN}",
"Accept": "application/vnd.github.v3+json",
"X-GitHub-Api-Version": "2022-11-28",
}
+15
View File
@@ -0,0 +1,15 @@
# Copyright 2025 Google LLC
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
from . import agent
+150
View File
@@ -0,0 +1,150 @@
# Copyright 2025 Google LLC
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# pylint: disable=g-importing-member
import os
from google.adk import Agent
import requests
GITHUB_TOKEN = os.getenv("GITHUB_TOKEN", "")
if not GITHUB_TOKEN:
raise ValueError("GITHUB_TOKEN environment variable not set")
OWNER = os.getenv("OWNER", "google")
REPO = os.getenv("REPO", "adk-python")
def get_github_pr_info_http(pr_number: int) -> str | None:
"""Fetches information for a GitHub Pull Request by sending direct HTTP requests.
Args:
pr_number (int): The number of the Pull Request.
Returns:
pr_message: A string.
"""
base_url = "https://api.github.com"
headers = {
"Accept": "application/vnd.github+json",
"Authorization": f"Bearer {GITHUB_TOKEN}",
"X-GitHub-Api-Version": "2022-11-28",
}
pr_message = ""
# --- 1. Get main PR details ---
pr_url = f"{base_url}/repos/{OWNER}/{REPO}/pulls/{pr_number}"
print(f"Fetching PR details from: {pr_url}")
try:
response = requests.get(pr_url, headers=headers)
response.raise_for_status()
pr_data = response.json()
pr_message += f"The PR title is: {pr_data.get('title')}\n"
except requests.exceptions.HTTPError as e:
print(
f"HTTP Error fetching PR details: {e.response.status_code} - "
f" {e.response.text}"
)
return None
except requests.exceptions.RequestException as e:
print(f"Network or request error fetching PR details: {e}")
return None
except Exception as e: # pylint: disable=broad-except
print(f"An unexpected error occurred: {e}")
return None
# --- 2. Fetching associated commits (paginated) ---
commits_url = pr_data.get(
"commits_url"
) # This URL is provided in the initial PR response
if commits_url:
print("\n--- Associated Commits in this PR: ---")
page = 1
while True:
# GitHub API often uses 'per_page' and 'page' for pagination
params = {
"per_page": 100,
"page": page,
} # Fetch up to 100 commits per page
try:
response = requests.get(commits_url, headers=headers, params=params)
response.raise_for_status()
commits_data = response.json()
if not commits_data: # No more commits
break
pr_message += "The associated commits are:\n"
for commit in commits_data:
message = commit.get("commit", {}).get("message", "").splitlines()[0]
if message:
pr_message += message + "\n"
# Check for 'Link' header to determine if more pages exist
# This is how GitHub's API indicates pagination
if "Link" in response.headers:
link_header = response.headers["Link"]
if 'rel="next"' in link_header:
page += 1 # Move to the next page
else:
break # No more pages
else:
break # No Link header, so probably only one page
except requests.exceptions.HTTPError as e:
print(
f"HTTP Error fetching PR commits (page {page}):"
f" {e.response.status_code} - {e.response.text}"
)
break
except requests.exceptions.RequestException as e:
print(
f"Network or request error fetching PR commits (page {page}): {e}"
)
break
else:
print("Commits URL not found in PR data.")
return pr_message
system_prompt = """
You are a helpful assistant to generate reasonable descriptions for pull requests for software engineers.
The descritions should not be too short (e.g.: less than 3 words), or too long (e.g.: more than 30 words).
The generated description should start with `chore`, `docs`, `feat`, `fix`, `test`, or `refactor`.
`feat` stands for a new feature.
`fix` stands for a bug fix.
`chore`, `docs`, `test`, and `refactor` stand for improvements.
Some good descriptions are:
1. feat: Added implementation for `get_eval_case`, `update_eval_case` and `delete_eval_case` for the local eval sets manager.
2. feat: Provide inject_session_state as public util method.
Some bad descriptions are:
1. fix: This fixes bugs.
2. feat: This is a new feature.
"""
root_agent = Agent(
model="gemini-2.0-flash",
name="github_pr_agent",
description="Generate pull request descriptions for ADK.",
instruction=system_prompt,
)
+73
View File
@@ -0,0 +1,73 @@
# Copyright 2025 Google LLC
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# pylint: disable=g-importing-member
import asyncio
import time
import agent
from google.adk.agents.run_config import RunConfig
from google.adk.runners import InMemoryRunner
from google.adk.sessions import Session
from google.genai import types
async def main():
app_name = "adk_pr_app"
user_id_1 = "adk_pr_user"
runner = InMemoryRunner(
agent=agent.root_agent,
app_name=app_name,
)
session_11 = await runner.session_service.create_session(
app_name=app_name, user_id=user_id_1
)
async def run_agent_prompt(session: Session, prompt_text: str):
content = types.Content(
role="user", parts=[types.Part.from_text(text=prompt_text)]
)
final_agent_response_parts = []
async for event in runner.run_async(
user_id=user_id_1,
session_id=session.id,
new_message=content,
run_config=RunConfig(save_input_blobs_as_artifacts=False),
):
if event.content.parts and event.content.parts[0].text:
if event.author == agent.root_agent.name:
final_agent_response_parts.append(event.content.parts[0].text)
print(f"<<<< Agent Final Output: {''.join(final_agent_response_parts)}\n")
pr_message = agent.get_github_pr_info_http(pr_number=1422)
query = "Generate pull request description for " + pr_message
await run_agent_prompt(session_11, query)
if __name__ == "__main__":
start_time = time.time()
print(
"Script start time:",
time.strftime("%Y-%m-%d %H:%M:%S", time.gmtime(start_time)),
)
print("------------------------------------")
asyncio.run(main())
end_time = time.time()
print("------------------------------------")
print(
"Script end time:",
time.strftime("%Y-%m-%d %H:%M:%S", time.gmtime(end_time)),
)
print("Total script execution time:", f"{end_time - start_time:.2f} seconds")
@@ -25,18 +25,18 @@ from adk_triaging_agent.utils import post_request
from google.adk import Agent
import requests
ALLOWED_LABELS = [
"documentation",
"services",
"question",
"tools",
"eval",
"live",
"models",
"tracing",
"core",
"web",
]
LABEL_TO_OWNER = {
"documentation": "polong",
"services": "DeanChensj",
"question": "",
"tools": "seanzhou1023",
"eval": "ankursharmas",
"live": "hangfei",
"models": "selcukgun",
"tracing": "Jacksunwei",
"core": "Jacksunwei",
"web": "wyf7107",
}
APPROVAL_INSTRUCTION = (
"Do not ask for user approval for labeling! If you can't find appropriate"
@@ -78,33 +78,61 @@ def list_unlabeled_issues(issue_count: int) -> dict[str, Any]:
return {"status": "success", "issues": unlabeled_issues}
def add_label_to_issue(issue_number: int, label: str) -> dict[str, Any]:
"""Add the specified label to the given issue number.
def add_label_and_owner_to_issue(
issue_number: int, label: str
) -> dict[str, Any]:
"""Add the specified label and owner to the given issue number.
Args:
issue_number: issue number of the Github issue.
label: label to assign
Returns:
The the status of this request, with the applied label when successful.
The the status of this request, with the applied label and assigned owner
when successful.
"""
print(f"Attempting to add label '{label}' to issue #{issue_number}")
if label not in ALLOWED_LABELS:
if label not in LABEL_TO_OWNER:
return error_response(
f"Error: Label '{label}' is not an allowed label. Will not apply."
)
url = f"{GITHUB_BASE_URL}/repos/{OWNER}/{REPO}/issues/{issue_number}/labels"
payload = [label, BOT_LABEL]
label_url = (
f"{GITHUB_BASE_URL}/repos/{OWNER}/{REPO}/issues/{issue_number}/labels"
)
label_payload = [label, BOT_LABEL]
try:
response = post_request(url, payload)
response = post_request(label_url, label_payload)
except requests.exceptions.RequestException as e:
return error_response(f"Error: {e}")
owner = LABEL_TO_OWNER.get(label, None)
if not owner:
return {
"status": "warning",
"message": (
f"{response}\n\nLabel '{label}' does not have an owner. Will not"
" assign."
),
"applied_label": label,
}
assignee_url = (
f"{GITHUB_BASE_URL}/repos/{OWNER}/{REPO}/issues/{issue_number}/assignees"
)
assignee_payload = {"assignees": [owner]}
try:
response = post_request(assignee_url, assignee_payload)
except requests.exceptions.RequestException as e:
return error_response(f"Error: {e}")
return {
"status": "success",
"message": response,
"applied_label": label,
"assigned_owner": owner,
}
@@ -128,9 +156,12 @@ root_agent = Agent(
- If it's agent orchestration, agent definition, label it with "core".
- If you can't find a appropriate labels for the issue, follow the previous instruction that starts with "IMPORTANT:".
Call the `add_label_and_owner_to_issue` tool to label the issue, which will also assign the issue to the owner of the label.
Present the followings in an easy to read format highlighting issue number and your label.
- the issue summary in a few sentence
- your label recommendation and justification
- the owner of the label if you assign the issue to an owner
""",
tools=[list_unlabeled_issues, add_label_to_issue],
tools=[list_unlabeled_issues, add_label_and_owner_to_issue],
)
@@ -14,10 +14,6 @@
from google.adk import Agent
from google.adk.tools import google_search
from google.genai import Client
# Only Vertex AI supports image generation for now.
client = Client()
root_agent = Agent(
model='gemini-2.0-flash-001',
+1
View File
@@ -51,6 +51,7 @@ dependencies = [
"typing-extensions>=4.5, <5",
"tzlocal>=5.3", # Time zone utilities
"uvicorn>=0.34.0", # ASGI server for FastAPI
"watchdog>=6.0.0", # For file change detection and hot reload
"websockets>=15.0.1", # For BaseLlmFlow
# go/keep-sorted end
]
+46
View File
@@ -0,0 +1,46 @@
# Copyright 2025 Google LLC
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
from __future__ import annotations
from typing import Union
from pydantic import RootModel
from ..utils.feature_decorator import working_in_progress
from .llm_agent import LlmAgentConfig
from .loop_agent import LoopAgentConfig
from .parallel_agent import ParallelAgentConfig
from .sequential_agent import SequentialAgentConfig
# A discriminated union of all possible agent configurations.
ConfigsUnion = Union[
LlmAgentConfig,
LoopAgentConfig,
ParallelAgentConfig,
SequentialAgentConfig,
]
# Use a RootModel to represent the agent directly at the top level.
# The `discriminator` is applied to the union within the RootModel.
@working_in_progress("AgentConfig is not ready for use.")
class AgentConfig(RootModel[ConfigsUnion]):
"""The config for the YAML schema to create an agent."""
class Config:
# Pydantic v2 requires this for discriminated unions on RootModel
# This tells the model to look at the 'agent_class' field of the input
# data to decide which model from the `ConfigsUnion` to use.
discriminator = "agent_class"
+187 -41
View File
@@ -19,9 +19,14 @@ from typing import Any
from typing import AsyncGenerator
from typing import Awaitable
from typing import Callable
from typing import Dict
from typing import final
from typing import Literal
from typing import Mapping
from typing import Optional
from typing import Type
from typing import TYPE_CHECKING
from typing import TypeVar
from typing import Union
from google.genai import types
@@ -34,6 +39,7 @@ from typing_extensions import override
from typing_extensions import TypeAlias
from ..events.event import Event
from ..utils.feature_decorator import working_in_progress
from .callback_context import CallbackContext
if TYPE_CHECKING:
@@ -56,6 +62,8 @@ AfterAgentCallback: TypeAlias = Union[
list[_SingleAgentCallback],
]
SelfAgent = TypeVar('SelfAgent', bound='BaseAgent')
class BaseAgent(BaseModel):
"""Base class for all agents in Agent Development Kit."""
@@ -121,6 +129,56 @@ class BaseAgent(BaseModel):
response and appended to event history as agent response.
"""
def clone(
self: SelfAgent, update: Mapping[str, Any] | None = None
) -> SelfAgent:
"""Creates a copy of this agent instance.
Args:
update: Optional mapping of new values for the fields of the cloned agent.
The keys of the mapping are the names of the fields to be updated, and
the values are the new values for those fields.
For example: {"name": "cloned_agent"}
Returns:
A new agent instance with identical configuration as the original
agent except for the fields specified in the update.
"""
if update is not None and 'parent_agent' in update:
raise ValueError(
'Cannot update `parent_agent` field in clone. Parent agent is set'
' only when the parent agent is instantiated with the sub-agents.'
)
# Only allow updating fields that are defined in the agent class.
allowed_fields = set(self.__class__.model_fields)
if update is not None:
invalid_fields = set(update) - allowed_fields
if invalid_fields:
raise ValueError(
f'Cannot update non-existent fields in {self.__class__.__name__}:'
f' {invalid_fields}'
)
cloned_agent = self.model_copy(update=update)
if update is None or 'sub_agents' not in update:
# If `sub_agents` is not provided in the update, need to recursively clone
# the sub-agents to avoid sharing the sub-agents with the original agent.
cloned_agent.sub_agents = []
for sub_agent in self.sub_agents:
cloned_sub_agent = sub_agent.clone()
cloned_sub_agent.parent_agent = cloned_agent
cloned_agent.sub_agents.append(cloned_sub_agent)
else:
for sub_agent in cloned_agent.sub_agents:
sub_agent.parent_agent = cloned_agent
# Remove the parent agent from the cloned agent to avoid sharing the parent
# agent with the cloned agent.
cloned_agent.parent_agent = None
return cloned_agent
@final
async def run_async(
self,
@@ -169,11 +227,18 @@ class BaseAgent(BaseModel):
"""
with tracer.start_as_current_span(f'agent_run [{self.name}]'):
ctx = self._create_invocation_context(parent_context)
# TODO(hangfei): support before/after_agent_callback
if event := await self.__handle_before_agent_callback(ctx):
yield event
if ctx.end_invocation:
return
async for event in self._run_live_impl(ctx):
yield event
if event := await self.__handle_after_agent_callback(ctx):
yield event
async def _run_async_impl(
self, ctx: InvocationContext
) -> AsyncGenerator[Event, None]:
@@ -277,73 +342,99 @@ class BaseAgent(BaseModel):
) -> Optional[Event]:
"""Runs the before_agent_callback if it exists.
Args:
ctx: InvocationContext, the invocation context for this agent.
Returns:
Optional[Event]: an event if callback provides content or changed state.
"""
ret_event = None
if not self.canonical_before_agent_callbacks:
return ret_event
callback_context = CallbackContext(ctx)
for callback in self.canonical_before_agent_callbacks:
before_agent_callback_content = callback(
callback_context=callback_context
)
if inspect.isawaitable(before_agent_callback_content):
before_agent_callback_content = await before_agent_callback_content
if before_agent_callback_content:
ret_event = Event(
invocation_id=ctx.invocation_id,
author=self.name,
branch=ctx.branch,
content=before_agent_callback_content,
actions=callback_context._event_actions,
# Run callbacks from the plugins.
before_agent_callback_content = (
await ctx.plugin_manager.run_before_agent_callback(
agent=self, callback_context=callback_context
)
ctx.end_invocation = True
return ret_event
)
# If no overrides are provided from the plugins, further run the canonical
# callbacks.
if (
not before_agent_callback_content
and self.canonical_before_agent_callbacks
):
for callback in self.canonical_before_agent_callbacks:
before_agent_callback_content = callback(
callback_context=callback_context
)
if inspect.isawaitable(before_agent_callback_content):
before_agent_callback_content = await before_agent_callback_content
if before_agent_callback_content:
break
# Process the override content if exists, and further process the state
# change if exists.
if before_agent_callback_content:
ret_event = Event(
invocation_id=ctx.invocation_id,
author=self.name,
branch=ctx.branch,
content=before_agent_callback_content,
actions=callback_context._event_actions,
)
ctx.end_invocation = True
return ret_event
if callback_context.state.has_delta():
ret_event = Event(
return Event(
invocation_id=ctx.invocation_id,
author=self.name,
branch=ctx.branch,
actions=callback_context._event_actions,
)
return ret_event
return None
async def __handle_after_agent_callback(
self, invocation_context: InvocationContext
) -> Optional[Event]:
"""Runs the after_agent_callback if it exists.
Args:
invocation_context: InvocationContext, the invocation context for this
agent.
Returns:
Optional[Event]: an event if callback provides content or changed state.
"""
ret_event = None
if not self.canonical_after_agent_callbacks:
return ret_event
callback_context = CallbackContext(invocation_context)
for callback in self.canonical_after_agent_callbacks:
after_agent_callback_content = callback(callback_context=callback_context)
if inspect.isawaitable(after_agent_callback_content):
after_agent_callback_content = await after_agent_callback_content
if after_agent_callback_content:
ret_event = Event(
invocation_id=invocation_context.invocation_id,
author=self.name,
branch=invocation_context.branch,
content=after_agent_callback_content,
actions=callback_context._event_actions,
# Run callbacks from the plugins.
after_agent_callback_content = (
await invocation_context.plugin_manager.run_after_agent_callback(
agent=self, callback_context=callback_context
)
return ret_event
)
if callback_context.state.has_delta():
# If no overrides are provided from the plugins, further run the canonical
# callbacks.
if (
not after_agent_callback_content
and self.canonical_after_agent_callbacks
):
for callback in self.canonical_after_agent_callbacks:
after_agent_callback_content = callback(
callback_context=callback_context
)
if inspect.isawaitable(after_agent_callback_content):
after_agent_callback_content = await after_agent_callback_content
if after_agent_callback_content:
break
# Process the override content if exists, and further process the state
# change if exists.
if after_agent_callback_content:
ret_event = Event(
invocation_id=invocation_context.invocation_id,
author=self.name,
@@ -351,8 +442,17 @@ class BaseAgent(BaseModel):
content=after_agent_callback_content,
actions=callback_context._event_actions,
)
return ret_event
return ret_event
if callback_context.state.has_delta():
return Event(
invocation_id=invocation_context.invocation_id,
author=self.name,
branch=invocation_context.branch,
content=after_agent_callback_content,
actions=callback_context._event_actions,
)
return None
@override
def model_post_init(self, __context: Any) -> None:
@@ -385,3 +485,49 @@ class BaseAgent(BaseModel):
)
sub_agent.parent_agent = self
return self
@classmethod
@working_in_progress('BaseAgent.from_config is not ready for use.')
def from_config(
cls: Type[SelfAgent],
config: BaseAgentConfig,
) -> SelfAgent:
"""Creates an agent from a config.
This method converts fields in a config to the corresponding
fields in an agent.
Child classes should re-implement this method to support loading from their
custom config types.
Args:
config: The config to create the agent from.
Returns:
The created agent.
"""
kwargs: Dict[str, Any] = {
'name': config.name,
'description': config.description,
}
return cls(**kwargs)
@working_in_progress('BaseAgentConfig is not ready for use.')
class BaseAgentConfig(BaseModel):
"""The config for the YAML schema of a BaseAgent.
Do not use this class directly. It's the base class for all agent configs.
"""
model_config = ConfigDict(extra='forbid')
agent_class: Literal['BaseAgent'] = 'BaseAgent'
"""Required. The class of the agent. The value is used to differentiate
among different agent classes."""
name: str
"""Required. The name of the agent."""
description: str = ''
"""Optional. The description of the agent."""

Some files were not shown because too many files have changed in this diff Show More