Repository Issue Activity (beta)

#6672 to_a2a() serves an agent card with capabilities.streaming=false while the mounted handler implements message/stream

#6871 to_openeval()/from_openeval(): interop between EvalSet and the EvalPort portable format

#7030 Artifact services (in_memory, file, GCS) key storage on raw session_id — padded id creates a session it cannot reach

#7029 scripts/check_new_py_files.py misses added files on Windows under jj and hg

#7027 [Bug] ADK 2.8.0: repeated HITL deadlocks on replay within one invocation (no LLM)

#7019 422 Unprocessable Entity in versions 2.6.0, 2.7.0, 2.8.0 when trying to save an eval case

#7024 Cloud telemetry export fails with invalid_scope when ADC is a service-account key file

#6966 feat(model_armor): screen tool output in ModelArmorPlugin (or companion plugin)

#7021 syria

#6748 Applications cannot influence what a toolset contributes when tool listing fails (per-user OAuth MCP servers)

#6941 Session id is normalized on write but not on read: after bfeb04c a padded id creates a session that cannot be read, deleted, or re-created (in_memory and sqlite, incl. the adk web / adk run store)

#6336 WEB UI Evals is not working

#7018 Instagram bot followers sabi ko dikhna chahiye

#7017 [BUG] BigQueryAgentAnalyticsPlugin re-runs all view DDL on every re-initialisation, on the request path (2.3.0 -> 2.8.0 regression)

#7014 Proposal: Add SnowflakeCortexAgent as a BaseAgent adapter

#6877 Annotated type hint metadata not preserved in function tool declarations/tool schema

#6998 VertexAiSessionService mutates local session before append_event succeeds

#7006 [Python][Live] Background tool output can reach the parent model during the transfer delay

#7010 `require_confirmation` callable returning a non-bool skips the confirmation gate: `cast(bool, ...)` is not a runtime check

#7009 chore: eval extra's google-cloud-aiplatform bound has no upper cap, unlike all/gcp/test

#7007 before_run early-exit events bypass on_event_callback

#7001 Vulnerability https://issuetracker.google.com/issues/556637640

#6584 RemoteA2aAgent: a terminal A2A task carrying tool activity fails is_final_response(), so the turn cannot be detected as finished (non-streaming peer)

#6954 [BUG]: LiteLlm.capabilities hardcodes output_schema_and_tools=True, bypassing the SetModelResponseTool workaround

#6948 Agent import validation can fail for newly created agent packages