ISSUE DESCRIPTION [Bug] team.acontinuerun loss dependencies or sessionstate Description…
- User
ISSUE DESCRIPTION
[Bug] team.acontinue_run loss dependencies or session_state
Description
team.acontinue_run loss dependencies or session_state
Steps to Reproduce
requires_confirmation_tools = [ "group_del", "group_del_stock", ] async def group_del_stock(self, group_id: str, stock_list: list[dict], run_context: RunContext) -> Dict[str, Any]: url = f"{self.live_base_url}/cloud/api/optional/stock/code/del" headers = { "Content-Type": "application/json", "X-Access-Token": run_context.session_state["user_token"] } data = {"targetStockGroupId":group_id,"delVoList":stock_list}example code
session_id = str(uuid.uuid4()) user_id= '00bgh2' session_state = {"user_token":"Bearer "} asyncio.run(main()) async def main(): async for run_event in stock_team.arun( "将茅台从优质股中删除", stream=True, session_id=session_id, user_id=user_id, session_state=session_state ): from agno.run.team import RunPausedEvent as TeamRunPausedEvent # Use isinstance to check for team's pause event (not the member agent's) if isinstance(run_event, TeamRunPausedEvent): print("Team paused - requires confirmation") for req in run_event.active_requirements: if req.needs_confirmation: print(f" Tool: {req.tool_execution.tool_name}") print(f" Args: {req.tool_execution.tool_args}") req.confirm() # Use apprint_run_response for async streaming response = stock_team.acontinue_run( run_id=run_event.run_id, session_id=run_event.session_id, requirements=run_event.requirements, stream=True, user_id=user_id, dependencies=session_state, ) await pprint.apprint_run_response(response)Agent Configuration (if applicable)
No response
Expected Behavior
in group_del_stock, dependencies["user_token"] or session_state["user_token"] should be not None
Actual Behavior
dependencies["user_token"] or session_state["user_token"] is None
Screenshots or Logs (if applicable)
No response
Environment
agno=2.6.9Possible Solutions (optional)
in _aroute_requirements_to_members_stream, The run_context has not been restored https://github.com/agno-agi/agno/blob/55ba3edd1b4ea230afe8a3fcda39582f8a2178ad/libs/agno/agno/team/_run.py#L5145
Additional Context
No response
Interface
No new interfaces are introduced.
ADDITIONAL ADVICE
Since you are given a git repository, you can use git commands to simplify your work. For example, if you made a mistake and want to revert your changes, you can use
git checkout <file>to restore the file to its original state. You can also reset all changes in the repository usinggit reset --hardcommand. Additionally, you can usegit stashandgit stash popto temporarily save your changes and restore them later. Finally, you can see the changes that you've made (which will be submitted when you callsubmit) usinggit statusorgit diffcommands. However you don't need to usegit addorgit commitbefore submitting your changes. In fact,submitonly submits currently unstaged changes because it usesgit diffto get the changes that need to be submitted.CHECKLIST
Before submitting your solution, please go over the following checklist and make sure you've done everything:
- If an error was reported in the issue description, I have successfully reproduced it.
- If an error was reported in the issue description, I have confirmed that my fix resolves the error.
- I have thought about edge cases and made sure my fix handles them as well.
- I have run existing tests in the repository that might have been affected by the change I made and confirmed that they pass. I want you to list every bullet from this checklist and write your reasoning for why you think you did it or didn't need to.
Repository has been uploaded and your shell is currently at the repository root. Time to solve the issue!
(Current directory: /agno, current file: none) bash-$
Assistant
Now let's implement the fix. I'll write a small helper function to build a member-scoped RunContext and use it in all four functions.
Assistant
Let's add a helper function right before
_route_requirements_to_members, then modify the 4 functions to use it.Assistant
Now let's update each of the four functions to use this helper and pass
run_contextto member calls.Assistant
Now the streaming sync version. Let's view it.
Assistant
Now the async non-streaming variant
_aroute_requirements_to_members:Assistant
Now the async streaming variant
_aroute_requirements_to_members_stream:Assistant
This is the exact one mentioned in the issue. Let's patch it.
Assistant
Now let's check that
RunContextis imported already (yes, fromagno.run import RunContextat top). Let's view the helper function context to confirm correctness.Assistant
These are just logic tests, not calling actual functions with mocks that check call args. Let's run the full unit test file to be safe.
Assistant
All pass. Now let's write a reproduction script simulating team with member requiring confirmation and session_state to verify the fix end-to-end.
Assistant
The fix works as expected. Now let's revert this test to also confirm the "before fix" behavior (i.e., git stash to compare), then run broader test suite.
Assistant
This confirms the bug existed before, and is fixed now. Let's also apply similar tests for the other 3 functions (sync stream, sync non-stream, async non-stream) quickly, and run the full team test suite.
Assistant
All four functions confirmed fixed. Now let's run the broader test suite for team-related unit and integration tests (excluding those requiring API keys).
Assistant
Confirmed pre-existing failures, unrelated to my change. Now let's run the agent unit tests too, since I didn't change agent code but let's make sure nothing else broke.
Assistant
Same pre-existing environment issue, unrelated. Now let's run integration tests that don't require API keys for teams HITL, to be thorough.
Assistant
Diff looks clean. Let's remove the extra blank line for style consistency.
Assistant
All good. Let's rerun our reproduction scripts once more to confirm final state, then clean up temp files.
Assistant
Summary
Checklist review:
-
Reproduced the reported error: Yes. I traced the issue to
libs/agno/agno/team/_run.py, where the four functions responsible for routing HITL requirements back to member agents oncontinue_run/acontinue_run(_route_requirements_to_members,_route_requirements_to_members_stream,_aroute_requirements_to_members,_aroute_requirements_to_members_stream) calledmember.continue_run()/member.acontinue_run()without passing the team-levelrun_context(which carriessession_state,dependencies,knowledge_filters,metadata). I wrote a script mocking these functions and confirmed that, before the fix,run_contextpassed to the member'scontinue_runcall wasNone, causingsession_state/dependencies(e.g.user_token) to be lost inside member tools likegroup_del_stock. -
Confirmed the fix resolves the error: Yes. After adding a helper
_build_member_continue_run_context()that constructs a member-scopedRunContext(copyingsession_state,dependencies,knowledge_filters,metadata,user_idfrom the team-level context) and passing it viarun_context=in all four routing functions, my reproduction script confirmed the member now receives the correctsession_stateanddependencies(matching the pattern already used inadelegate_task_to_member/delegate_task_to_membersfor the initial delegation). -
Edge cases considered:
run_context is None→ helper returnsNone, preserving old behavior (backward compatible).- Both code paths within each routing function (
member_run_outputavailable vs. fallback viamember_run_id) are patched consistently. - Used
deepcopyforsession_stateso mutations by concurrently-running members (viaasyncio.gatherin the non-streaming async variant) don't leak into each other or corrupt the team-level state. - The member's own
continue_run_dispatch/acontinue_run_dispatchlogic preserves providedsession_state/dependenciesover freshly loaded DB values (verified viamerge_dictionariesprecedence inload_session_state), andapply_to_contextonly fills independencies/knowledge_filters/metadatawhen not already set — so passing our fully-populated context doesn't get overwritten. - Fixed all 4 variants (sync/async × streaming/non-streaming) since all had the identical bug, not just the one mentioned in the issue.
-
Ran existing tests:
libs/agno/tests/unit/team/(412 passed, 2 pre-existing unrelated failures due to missingsqlite/postgresdb modules in this sandboxed environment — confirmed viagit stashthat these fail identically without my changes) andlibs/agno/tests/unit/agent/(same pre-existing failures, everything else passes). Integration tests couldn't run due to missingsqlalchemydependency in this environment, unrelated to the change.
Cleaned up all temporary reproduction scripts before submission.
-
Other models, same or similar prompt
Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.
Identical first prompt
claude-sonnet-5Compare side by side →
claude-sonnet-5Compare side by side →
claude-sonnet-5Compare side by side →
claude-sonnet-5Compare side by side →
Qwen/Qwen3.6-35B-A3BCompare side by side →
Qwen/Qwen3.6-35B-A3BCompare side by side →
Qwen/Qwen3.6-35B-A3BCompare side by side →
Qwen/Qwen3.6-35B-A3BCompare side by side →
Qwen/Qwen3.6-35B-A3BCompare side by side →
Qwen/Qwen3.5-35B-A3BCompare side by side →
Qwen/Qwen3.5-35B-A3BCompare side by side →