[pull] main from google:main#401
Merged
Merged
Conversation
The instruction told the model to use a `log_query` tool that is not registered on the agent, so when the model followed it the run failed with "Tool 'log_query' not found". That made the integration_test presubmit, which runs this sample, flaky. Reword the instruction to just reply, matching the sample's documented output. Co-authored-by: George Weale <gweale@google.com> PiperOrigin-RevId: 943350141
Merge #5523 Closes #5486 ## Summary This PR fixes Windows handling of local `file://` artifact URIs. Both `google/adk/cli/service_registry.py` and `google/adk/artifacts/file_artifact_service.py` were passing URI-style paths into `Path(...)` without Windows URI-to-filesystem normalization. On Windows, this can produce malformed drive-relative paths such as `C:foo` instead of absolute paths like `C:\foo`. This change applies `url2pathname()` under `os.name == "nt"` in both locations so canonical file URIs like `file:///C:/...` are converted to proper native Windows paths before constructing `Path(...)`. ## Validation Manually tested on Windows with: 1. current working directory on `C:` and artifact target on `C:` - before the patch, the bug reproduced and the artifact root was created in the wrong location - after the patch, the artifact root was created in the intended location 2. current working directory on `C:` and artifact target on `D:` - before the patch, the current code could appear to work - after the patch, it still worked, now through correct URI-to-path normalization 3. paths containing spaces - after the patch, the artifact root was created correctly in the intended directory I also verified the drive-relative behavior separately with: ```python import ntpath print(ntpath.abspath("C:foo")) print(ntpath.abspath("D:foo")) ``` With current working directory `C:\Users\user1\projects\agent1`, this produced: ```text C:\Users\user1\projects\agent1\foo D:\foo ``` This matches the observed issue: without Windows-specific normalization, the malformed URI-derived path can behave like a drive-relative path such as `C:foo` instead of a proper absolute path like `C:\foo`. ## Notes I have not added a regression test in this PR yet. I can add one in a follow-up if maintainers prefer a specific test location for this path-conversion logic. Co-authored-by: George Weale <gweale@google.com> COPYBARA_INTEGRATE_REVIEW=#5523 from cyrus-mz:fix/windows-file-artifact-uri e93847b PiperOrigin-RevId: 943367471
…play synchronization Extract `ReplayManager` to unify session event scanning, replay interception, and sequence barrier management across static and dynamic workflow nodes. Co-authored-by: Shangjie Chen <deanchen@google.com> PiperOrigin-RevId: 943376672
Merge #6314 PiperOrigin-RevId: 943391685
The userinfo_endpoint field is never read anywhere in ADK, so this hardcoded URL had no runtime effect. The OAuth flow only uses the authorization and token endpoints. Removing it. Co-authored-by: George Weale <gweale@google.com> PiperOrigin-RevId: 943396380
Merge #4271 **Please ensure you have read the [contribution guide](https://github.com/google/adk-python/blob/main/CONTRIBUTING.md) before creating a pull request.** ### Link to Issue or Description of Change - Closes: #4270 **Problem:** When using MCP tools with McpToolset, configuration information for the object is not exposed as object properties or attributes meaning that 'private' attributes need to be accessed, e.g. toolset._connection_params rather than toolset.connection_params **Solution:** Exposed them as properties ### Testing Plan **Unit Tests:** - [x] I have added or updated unit tests for my change. - [x] All unit tests pass locally. **Manual End-to-End (E2E) Tests:** I have added new test cases also checked locally ### Checklist - [x] I have read the [CONTRIBUTING.md](https://github.com/google/adk-python/blob/main/CONTRIBUTING.md) document. - [x] I have performed a self-review of my own code. - [x] I have commented my code, particularly in hard-to-understand areas. - [x] I have added tests that prove my fix is effective or that my feature works. - [x] New and existing unit tests pass locally with my changes. - [x] I have manually tested my changes end-to-end. - [x] Any dependent changes have been merged and published in downstream modules. ### Additional context Co-authored-by: Haran Rajkumar <haranrk@google.com> COPYBARA_INTEGRATE_REVIEW=#4271 from DineshThumma9:mcp-missing-attr 3d48401 PiperOrigin-RevId: 943396972
Add ManagedAgent(BaseAgent), which drives the Managed Agents interactions.create API directly. This first cut supports server-side tools only (ADK built-in tools and raw types.Tool configs); client-executed tools (FunctionTool/callables) and MCP are rejected with NotImplementedError. Multi-turn chaining reuses previous_interaction_id and recovers the sandbox environment across turns. Co-authored-by: Haran Rajkumar <haranrk@google.com> PiperOrigin-RevId: 943406685
Introduce `append` method and `create_sub_branch` classmethod on `_BranchPath` to standardize dynamic branch path creation across workflow node runners, agent tools, and parallel agent execution. Co-authored-by: Shangjie Chen <deanchen@google.com> PiperOrigin-RevId: 943430872
Merge #5991 # Add Transient Model Input Context to RunConfig ### Link to Issue or Description of Change **1. Link to an existing issue (if applicable):** - Closes: #5990 **Problem:** Host applications sometimes need to provide request-scoped context to an agent for a single invocation without persisting that context into `session.events`. Before this change, callers had to either append synthetic session events or merge application context into the user message. Both approaches blur the boundary between durable conversation history and transient model input. **Solution:** This change adds `RunConfig.model_input_context`, a list of `google.genai.types.Content` values that are injected into the LLM request for the current invocation only. The context is deep-copied before insertion, added before the invocation user content, and never appended to `session.events`. The insertion path is separate from instruction-related content so the transient context keeps a stable position across tool-call loops and multi-agent `include_contents="none"` flows. ### Testing Plan **Unit Tests:** - [x] I have added or updated unit tests for my change. - [x] All unit tests pass locally. Passed locally: ```text .venv/bin/python -m pytest tests/unittests/agents/test_llm_agent_include_contents.py tests/unittests/agents/test_run_config.py 14 passed ``` Additional checks: ```text .venv/bin/pyink --check --diff --config pyproject.toml src/google/adk/flows/llm_flows/contents.py tests/unittests/agents/test_llm_agent_include_contents.py tests/unittests/agents/test_run_config.py .venv/bin/isort --check-only src/google/adk/flows/llm_flows/contents.py tests/unittests/agents/test_llm_agent_include_contents.py tests/unittests/agents/test_run_config.py src/google/adk/agents/run_config.py git diff --check ``` **Manual End-to-End (E2E) Tests:** Not run. This change is covered by focused unit tests around LLM request construction and session persistence. ### Checklist - [x] I have read the [CONTRIBUTING.md](https://github.com/google/adk-python/blob/main/CONTRIBUTING.md) document. - [x] I have performed a self-review of my own code. - [x] I have commented my code, particularly in hard-to-understand areas. - [x] I have added tests that prove my fix is effective or that my feature works. - [x] New and existing unit tests pass locally with my changes. - [x] I have manually tested my changes end-to-end. - [ ] Any dependent changes have been merged and published in downstream modules. ### Additional context The unit coverage verifies that transient context: - is sent to the model without being persisted to the session; - stays before the invocation user message after a tool call; - works for a sub-agent using `include_contents="none"` in a sequential flow; - is accepted by `RunConfig` as `types.Content`. Co-authored-by: George Weale <gweale@google.com> COPYBARA_INTEGRATE_REVIEW=#5991 from nicolasmota:spike/model-input-context c835972 PiperOrigin-RevId: 943484488
…cs views Add usage_thinking_tokens and usage_tool_use_tokens to the LLM_RESPONSE analytics view, sourced from the usage_metadata proto the plugin already logs to attributes.usage_metadata (thoughts_token_count and tool_use_prompt_token_count). Per the genai GenerateContentResponseUsageMetadata contract, total_token_count = prompt_token_count + candidates_token_count + tool_use_prompt_token_count + thoughts_token_count, so thinking and tool-use tokens are separate addends rather than subsets of prompt/candidates. Surfacing them lets analytics account for them without double counting. Both fields are optional and resolve to NULL for models/responses that do not report them, so the change stays model-agnostic. No plugin logging change is needed since the full usage_metadata is already persisted to attributes. Co-authored-by: Haiyuan Cao <haiyuan@google.com> PiperOrigin-RevId: 943493796
Introduce NodeTool, allowing individual Nodes and entire Workflows to be wrapped and executed as standard ADK Tools. This PR implements the core, single-turn execution capability: - Support auto-wrapping of BaseNode (Workflows) directly in Agent.tools. - Wrapping synchronous and asynchronous function nodes as NodeTools. - Providing a complete sample workflow demonstrating how to run a workflow as a tool. Resumption and multi-turn nested HITL support are skipped in this PR and will be fully enabled in the later PR. Co-authored-by: Shangjie Chen <deanchen@google.com> PiperOrigin-RevId: 943499058
PiperOrigin-RevId: 943500722
To prevent OOM issues if HTTP request/response bodies are too long Co-authored-by: Kathy Wu <wukathy@google.com> PiperOrigin-RevId: 943502772
Add scripts/check_new_py_files.py, which enforces ADK's private-by-default policy: a newly-added Python file under src/google/adk/ must have a '_'-prefixed basename (expose public API via __init__.py / __all__ instead). Newly-added files are detected by comparing the checked-out tree against a baseline source tree, so the check needs no git history. Co-authored-by: Haran Rajkumar <haranrk@google.com> PiperOrigin-RevId: 943513793
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
11 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
See Commits and Changes for more details.
Created by
pull[bot] (v2.0.0-alpha.4)
Can you help keep this open source service alive? 💖 Please sponsor : )