Merge branch 'main' into use-starred-tuple-unpacking-on-gcs-artifact-blob-names-versions

Author: zshehov

.gemini/settings.json

@@ -0,0 +1,3 @@
+{
+  "contextFileName": "AGENTS.md"
+}

.github/workflows/pr-commit-check.yml

@@ -0,0 +1,62 @@
+# .github/workflows/pr-commit-check.yml
+# This GitHub Action workflow checks if a pull request has more than one commit.
+# If it does, it fails the check and instructs the user to squash their commits.
+
+name: 'PR Commit Check'
+
+# This workflow runs on pull request events.
+# It's configured to run on any pull request that is opened or synchronized (new commits pushed).
+on:
+  pull_request:
+    types: [opened, synchronize]
+
+# Defines the jobs that will run as part of the workflow.
+jobs:
+  check-commit-count:
+    # The type of runner that the job will run on. 'ubuntu-latest' is a good default.
+    runs-on: ubuntu-latest
+
+    # The steps that will be executed as part of the job.
+    steps:
+      # Step 1: Check out the code
+      # This action checks out your repository under $GITHUB_WORKSPACE, so your workflow can access it.
+      - name: Checkout Code
+        uses: actions/checkout@v4
+        with:
+          # We need to fetch all commits to accurately count them.
+          # '0' means fetch all history for all branches and tags.
+          fetch-depth: 0
+
+      # Step 2: Count the commits in the pull request
+      # This step runs a script to get the number of commits in the PR.
+      - name: Count Commits
+        id: count_commits
+        # We use `git rev-list --count` to count the commits.
+        # ${{ github.event.pull_request.base.sha }} is the commit SHA of the base branch.
+        # ${{ github.event.pull_request.head.sha }} is the commit SHA of the head branch (the PR branch).
+        # The '..' syntax gives us the list of commits in the head branch that are not in the base branch.
+        # The output of the command (the count) is stored in a step output variable named 'count'.
+        run: |
+          count=$(git rev-list --count ${{ github.event.pull_request.base.sha }}..${{ github.event.pull_request.head.sha }})
+          echo "commit_count=$count" >> $GITHUB_OUTPUT
+
+      # Step 3: Check if the commit count is greater than 1
+      # This step uses the output from the previous step to decide whether to pass or fail.
+      - name: Check Commit Count
+        # This step only runs if the 'commit_count' output from the 'count_commits' step is greater than 1.
+        if: steps.count_commits.outputs.commit_count > 1
+        # If the condition is met, the workflow will exit with a failure status.
+        run: |
+          echo "This pull request has ${{ steps.count_commits.outputs.commit_count }} commits."
+          echo "Please squash them into a single commit before merging."
+          echo "You can use git rebase -i HEAD~N"
+          echo "...where N is the number of commits you want to squash together. The PR check conveniently tells you this number! For example, if the check says you have 3 commits, you would run: git rebase -i HEAD~3."
+          echo "Because you have rewritten the commit history, you must use the --force flag to update the pull request: git push --force"
+          exit 1
+
+      # Step 4: Success message
+      # This step runs if the commit count is not greater than 1 (i.e., it's 1).
+      - name: Success
+        if: steps.count_commits.outputs.commit_count <= 1
+        run: |
+          echo "This pull request has a single commit. Great job!"

.github/workflows/triage.yml

@@ -40,4 +40,5 @@ jobs:
           ISSUE_TITLE: ${{ github.event.issue.title }}
           ISSUE_BODY: ${{ github.event.issue.body }}
           ISSUE_COUNT_TO_PROCESS: '3' # Process 3 issues at a time on schedule
-        run: python contributing/samples/adk_triaging_agent/main.py
+          PYTHONPATH: contributing/samples
+        run: python -m adk_triaging_agent.main

AGENTS.md

@@ -0,0 +1,114 @@
+# Gemini CLI / Gemini Code Assist Context
+
+This document provides context for the Gemini CLI and Gemini Code Assist to understand the project and assist with development.
+
+## Project Overview
+
+The Agent Development Kit (ADK) is an open-source, code-first Python toolkit for building, evaluating, and deploying sophisticated AI agents with flexibility and control. While optimized for Gemini and the Google ecosystem, ADK is model-agnostic, deployment-agnostic, and is built for compatibility with other frameworks. ADK was designed to make agent development feel more like software development, to make it easier for developers to create, deploy, and orchestrate agentic architectures that range from simple tasks to complex workflows.
+
+## ADK: Style Guides
+
+### Python Style Guide
+
+The project follows the Google Python Style Guide. Key conventions are enforced using `pylint` with the provided `pylintrc` configuration file. Here are some of the key style points:
+
+*   **Indentation**: 2 spaces.
+*   **Line Length**: Maximum 80 characters.
+*   **Naming Conventions**:
+    *   `function_and_variable_names`: `snake_case`
+    *   `ClassNames`: `CamelCase`
+    *   `CONSTANTS`: `UPPERCASE_SNAKE_CASE`
+*   **Docstrings**: Required for all public modules, functions, classes, and methods.
+*   **Imports**: Organized and sorted.
+*   **Error Handling**: Specific exceptions should be caught, not general ones like `Exception`.
+
+### Autoformat
+
+We have autoformat.sh to help solve import organize and formatting issues.
+
+```bash
+# Run in open_source_workspace/
+$ ./autoformat.sh
+```
+
+### In ADK source
+
+Below styles applies to the ADK source code (under `src/` folder of the Github.
+repo).
+
+#### Use relative imports
+
+```python
+# DO
+from ..agents.llm_agent import LlmAgent
+
+# DON'T
+from google.adk.agents.llm_agent import LlmAgent
+```
+
+#### Import from module, not from `__init__.py`
+
+```python
+# DO
+from ..agents.llm_agent import LlmAgent
+
+# DON'T
+from ..agents  import LlmAgent # import from agents/__init__.py
+```
+
+#### Always do `from __future__ import annotations`
+
+```python
+# DO THIS, right after the open-source header.
+from __future__ import annotations
+```
+
+Like below:
+
+```python
+# 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
+
+# ... the rest of the file.
+```
+
+This allows us to forward-reference a class without quotes.
+
+Check out go/pep563 for details.
+
+### In ADK tests
+
+#### Use absolute imports
+
+In tests, we use `google.adk` same as how our users uses.
+
+```python
+# DO
+from google.adk.agents.llm_agent import LlmAgent
+
+# DON'T
+from ..agents.llm_agent import LlmAgent
+```
+
+## ADK: Local testing
+
+### Unit tests
+
+Run below command:
+
+```bash
+$ pytest tests/unittests
+```

CONTRIBUTING.md

@@ -49,6 +49,7 @@ This project follows
 
 ## Requirement for PRs
 
+- Each PR should only have one commit. Please squash it if there are multiple PRs.
 - All PRs, other than small documentation or typo fixes, should have a Issue assoicated. If not, please create one.
 - Small, focused PRs. Keep changes minimal—one concern per PR.
 - For bug fixes or features, please provide logs or screenshot after the fix is applied to help reviewers better understand the fix.
@@ -147,11 +148,11 @@ For any changes that impact user-facing documentation (guides, API reference, tu
     pytest ./tests/unittests
     ```
 
-    NOTE: for accurately repro test failure, only include `test` and `eval` as
-    extra dependencies.
+    NOTE: for accurate repro of test failure, only include `test`, `eval` and 
+    `a2a` as extra dependencies.
 
     ```shell
-    uv sync --extra test --extra eval
+    uv sync --extra test --extra eval --extra a2a
     pytest ./tests/unittests
     ```
 

contributing/samples/a2a_auth/README.md

@@ -0,0 +1,183 @@
+# A2A OAuth Authentication Sample Agent
+
+This sample demonstrates the **Agent-to-Agent (A2A)** architecture with **OAuth Authentication** workflows in the Agent Development Kit (ADK). The sample implements a multi-agent system where a remote agent can surface OAuth authentication requests to the local agent, which then guides the end user through the OAuth flow before returning the authentication credentials to the remote agent for API access.
+
+## Overview
+
+The A2A OAuth Authentication sample consists of:
+
+- **Root Agent** (`root_agent`): The main orchestrator that handles user requests and delegates tasks to specialized agents
+- **YouTube Search Agent** (`youtube_search_agent`): A local agent that handles YouTube video searches using LangChain tools
+- **BigQuery Agent** (`bigquery_agent`): A remote A2A agent that manages BigQuery operations and requires OAuth authentication for Google Cloud access
+
+## Architecture
+
+```
+┌─────────────────┐    ┌────────────────────┐    ┌──────────────────┐
+│   End User      │───▶│   Root Agent       │───▶│   BigQuery Agent │
+│   (OAuth Flow)  │    │    (Local)         │    │  (Remote A2A)    │
+│                 │    │                    │    │ (localhost:8001) │
+│   OAuth UI      │◀───│                    │◀───│   OAuth Request  │
+└─────────────────┘    └────────────────────┘    └──────────────────┘
+```
+
+## Key Features
+
+### 1. **Multi-Agent Architecture**
+- Root agent coordinates between local YouTube search and remote BigQuery operations
+- Demonstrates hybrid local/remote agent workflows
+- Seamless task delegation based on user request types
+
+### 2. **OAuth Authentication Workflow**
+- Remote BigQuery agent surfaces OAuth authentication requests to the root agent
+- Root agent guides end users through Google OAuth flow for BigQuery access
+- Secure token exchange between agents for authenticated API calls
+
+### 3. **Google Cloud Integration**
+- BigQuery toolset with comprehensive dataset and table management capabilities
+- OAuth-protected access to user's Google Cloud BigQuery resources
+- Support for listing, creating, and managing datasets and tables
+
+### 4. **LangChain Tool Integration**
+- YouTube search functionality using LangChain community tools
+- Demonstrates integration of third-party tools in agent workflows
+
+## Setup and Usage
+
+### Prerequisites
+
+1. **Set up OAuth Credentials**:
+   ```bash
+   export OAUTH_CLIENT_ID=your_google_oauth_client_id
+   export OAUTH_CLIENT_SECRET=your_google_oauth_client_secret
+   ```
+
+2. **Start the Remote BigQuery Agent server**:
+   ```bash
+   # Start the remote a2a server that serves the BigQuery agent on port 8001
+   adk api_server --a2a --port 8001 contributing/samples/a2a_auth/remote_a2a
+   ```
+
+3. **Run the Main Agent**:
+   ```bash
+   # In a separate terminal, run the adk web server
+   adk web contributing/samples/
+   ```
+
+### Example Interactions
+
+Once both services are running, you can interact with the root agent:
+
+**YouTube Search (No Authentication Required):**
+```
+User: Search for 3 Taylor Swift music videos
+Agent: I'll help you search for Taylor Swift music videos on YouTube.
+[Agent delegates to YouTube Search Agent]
+Agent: I found 3 Taylor Swift music videos:
+1. "Anti-Hero" - Official Music Video
+2. "Shake It Off" - Official Music Video
+3. "Blank Space" - Official Music Video
+```
+
+**BigQuery Operations (OAuth Required):**
+```
+User: List my BigQuery datasets
+Agent: I'll help you access your BigQuery datasets. This requires authentication with your Google account.
+[Agent delegates to BigQuery Agent]
+Agent: To access your BigQuery data, please complete the OAuth authentication.
+[OAuth flow initiated - user redirected to Google authentication]
+User: [Completes OAuth flow in browser]
+Agent: Authentication successful! Here are your BigQuery datasets:
+- dataset_1: Customer Analytics
+- dataset_2: Sales Data
+- dataset_3: Marketing Metrics
+```
+
+**Dataset Management:**
+```
+User: Show me details for my Customer Analytics dataset
+Agent: I'll get the details for your Customer Analytics dataset.
+[Using existing OAuth token]
+Agent: Customer Analytics Dataset Details:
+- Created: 2024-01-15
+- Location: US
+- Tables: 5
+- Description: Customer behavior and analytics data
+```
+
+## Code Structure
+
+### Main Agent (`agent.py`)
+
+- **`youtube_search_agent`**: Local agent with LangChain YouTube search tool
+- **`bigquery_agent`**: Remote A2A agent configuration for BigQuery operations
+- **`root_agent`**: Main orchestrator with task delegation logic
+
+### Remote BigQuery Agent (`remote_a2a/bigquery_agent/`)
+
+- **`agent.py`**: Implementation of the BigQuery agent with OAuth toolset
+- **`agent.json`**: Agent card of the A2A agent
+- **`BigQueryToolset`**: OAuth-enabled tools for BigQuery dataset and table management
+
+## OAuth Authentication Workflow
+
+The OAuth authentication process follows this pattern:
+
+1. **Initial Request**: User requests BigQuery operation through root agent
+2. **Delegation**: Root agent delegates to remote BigQuery agent
+3. **Auth Check**: BigQuery agent checks for valid OAuth token
+4. **Auth Request**: If no token, agent surfaces OAuth request to root agent
+5. **User OAuth**: Root agent guides user through Google OAuth flow
+6. **Token Exchange**: Root agent sends OAuth token to BigQuery agent
+7. **API Call**: BigQuery agent uses token to make authenticated API calls
+8. **Result Return**: BigQuery agent returns results through root agent to user
+
+## Supported BigQuery Operations
+
+The BigQuery agent supports the following operations:
+
+### Dataset Operations:
+- **List Datasets**: `bigquery_datasets_list` - Get all user's datasets
+- **Get Dataset**: `bigquery_datasets_get` - Get specific dataset details
+- **Create Dataset**: `bigquery_datasets_insert` - Create new dataset
+
+### Table Operations:
+- **List Tables**: `bigquery_tables_list` - Get tables in a dataset
+- **Get Table**: `bigquery_tables_get` - Get specific table details
+- **Create Table**: `bigquery_tables_insert` - Create new table in dataset
+
+## Extending the Sample
+
+You can extend this sample by:
+
+- Adding more Google Cloud services (Cloud Storage, Compute Engine, etc.)
+- Implementing token refresh and expiration handling
+- Adding role-based access control for different BigQuery operations
+- Creating OAuth flows for other providers (Microsoft, Facebook, etc.)
+- Adding audit logging for authentication events
+- Implementing multi-tenant OAuth token management
+
+## Troubleshooting
+
+**Connection Issues:**
+- Ensure the local ADK web server is running on port 8000
+- Ensure the remote A2A server is running on port 8001
+- Check that no firewall is blocking localhost connections
+- Verify the agent.json URL matches the running A2A server
+
+**OAuth Issues:**
+- Verify OAuth client ID and secret are correctly set in .env file
+- Ensure OAuth redirect URIs are properly configured in Google Cloud Console
+- Check that the OAuth scopes include BigQuery access permissions
+- Verify the user has access to the BigQuery projects/datasets
+
+**BigQuery Access Issues:**
+- Ensure the authenticated user has BigQuery permissions
+- Check that the Google Cloud project has BigQuery API enabled
+- Verify dataset and table names are correct and accessible
+- Check for quota limits on BigQuery API calls
+
+**Agent Communication Issues:**
+- Check the logs for both the local ADK web server and remote A2A server
+- Verify OAuth tokens are properly passed between agents
+- Ensure agent instructions are clear about authentication requirements