adk-python · Issues· 494 open

[Bug]: LocalEvalSampler treats explicit empty eval case ID lists as all cases

ApplicationIntegrationToolset silently builds no tools when a Connectors schema operation fails, and waits forever on one that never finishes

RestApiTool: integer / boolean / array header and cookie parameters crash the agent run (httpx: Header value must be str or bytes)

RestApiTool writes the tool's API key / OAuth token into the caller's args, so it lands in trace spans and after-tool callbacks

BigQuery execute_sql blocks the event loop for the whole query

hallucinations_v1: sentence validator ignores grounding_metadata

to_mcp_server never deletes the ADK session of a dead MCP connection; a stateless streamable HTTP deployment leaks one session per tool call

AutoTracingPlugin rebinds what getmembers() returns, not what the class holds: @staticmethod becomes an instance method, @classmethod is never traced, base methods get pinned onto subclasses

Memory has no removal path: delete_session() leaves what add_session_to_memory() copied, and BaseMemoryService exposes no way to remove it

A2A message:stream against a deployed Agent Engine (reasoningEngines) target truncates after TASK_STATE_WORKING, or fails with 'append=True for nonexistent artifact_id'

GCPSkillRegistry rejects every Google-published skill: dotted ids (cloud.google.com-*) fail name validation in get_skill and search_skills

RemoteA2aAgent enforces HTTPS for agent cards, blocking plaintext A2A inside a service mesh (mTLS)

LLM-as-judge default config triggers google-genai's AFC warning in every eval run

ToolConfirmation deny verdict is never enforced by the framework: custom BaseTool executes after user clicks Decline (adk-python 2.9.1)

set_model_response is missing from the MCP _RESERVED_TOOL_NAMES guard