feat(delegation): background fan-out — parallel subagents, one consolidated return - #49734
Merged
Conversation
Contributor
🔎 Lint report:
|
| Rule | Count |
|---|---|
invalid-argument-type |
2 |
First entries
tests/tools/test_async_delegation.py:421: [invalid-argument-type] invalid-argument-type: Argument to function `AIAgent._dispatch_delegate_task` is incorrect: Expected `AIAgent`, found `_FakeAgent`
tools/delegate_tool.py:2526: [invalid-argument-type] invalid-argument-type: Argument to function `dispatch_async_delegation_batch` is incorrect: Expected `list[str]`, found `list[Any | str | None | list[str]]`
✅ Fixed issues (10):
| Rule | Count |
|---|---|
invalid-argument-type |
10 |
First entries
tools/delegate_tool.py:2470: [invalid-argument-type] invalid-argument-type: Method `__getitem__` of type `Overload[(i: SupportsIndex, /) -> dict[str, str | None | list[str]], (s: slice[SupportsIndex | None, SupportsIndex | None, SupportsIndex | None], /) -> list[dict[str, str | None | list[str]]]]` cannot be called with key of type `str` on object of type `list[dict[str, str | None | list[str]]]`
tools/delegate_tool.py:2470: [invalid-argument-type] invalid-argument-type: Method `__getitem__` of type `Overload[(i: SupportsIndex, /) -> dict[str, str | None | list[str]], (s: slice[SupportsIndex | None, SupportsIndex | None, SupportsIndex | None], /) -> list[dict[str, str | None | list[str]]]]` cannot be called with key of type `None` on object of type `list[dict[str, str | None | list[str]]]`
tools/delegate_tool.py:2470: [invalid-argument-type] invalid-argument-type: Method `__getitem__` of type `Overload[(i: SupportsIndex, /) -> dict[str, Any], (s: slice[SupportsIndex | None, SupportsIndex | None, SupportsIndex | None], /) -> list[dict[str, Any]]]` cannot be called with key of type `None` on object of type `list[dict[str, Any]]`
tools/delegate_tool.py:2470: [invalid-argument-type] invalid-argument-type: Method `__getitem__` of type `Overload[(i: SupportsIndex, /) -> dict[str, Any], (s: slice[SupportsIndex | None, SupportsIndex | None, SupportsIndex | None], /) -> list[dict[str, Any]]]` cannot be called with key of type `str` on object of type `list[dict[str, Any]]`
tools/delegate_tool.py:2306: [invalid-argument-type] invalid-argument-type: Argument to function `dispatch_async_delegation` is incorrect: Expected `str`, found `Any | str | None | list[str]`
tools/delegate_tool.py:2478: [invalid-argument-type] invalid-argument-type: Method `__getitem__` of type `Overload[(i: SupportsIndex, /) -> tuple[int, dict[str, Any] | dict[str, str | None | list[str]], Unknown], (s: slice[SupportsIndex | None, SupportsIndex | None, SupportsIndex | None], /) -> list[tuple[int, dict[str, Any] | dict[str, str | None | list[str]], Unknown]]]` cannot be called with key of type `str` on object of type `list[tuple[int, dict[str, Any] | dict[str, str | None | list[str]], Unknown]]`
tools/delegate_tool.py:2478: [invalid-argument-type] invalid-argument-type: Method `__getitem__` of type `Overload[(i: SupportsIndex, /) -> tuple[int, dict[str, Any] | dict[str, str | None | list[str]], Unknown], (s: slice[SupportsIndex | None, SupportsIndex | None, SupportsIndex | None], /) -> list[tuple[int, dict[str, Any] | dict[str, str | None | list[str]], Unknown]]]` cannot be called with key of type `None` on object of type `list[tuple[int, dict[str, Any] | dict[str, str | None | list[str]], Unknown]]`
tools/delegate_tool.py:2307: [invalid-argument-type] invalid-argument-type: Argument to function `dispatch_async_delegation` is incorrect: Expected `str | None`, found `Any | None | str | list[str]`
tools/delegate_tool.py:2294: [invalid-argument-type] invalid-argument-type: Argument to function `_run_single_child` is incorrect: Expected `str`, found `Unknown | str | None | list[str]`
tools/delegate_tool.py:2308: [invalid-argument-type] invalid-argument-type: Argument to function `dispatch_async_delegation` is incorrect: Expected `list[str] | None`, found `(Any & ~AlwaysFalsy) | (str & ~AlwaysFalsy) | list[str] | None`
Unchanged: 5907 pre-existing issues carried over.
Diagnostics are surfaced as warnings — this check never fails the build.
alt-glitch
added
type/feature
teknium1
force-pushed
the
hermes/hermes-572aa955
branch
from
June 20, 2026 17:49
20cb760 to
b897cce
Compare
teknium1
changed the title
…ound The model no longer decides whether a subagent runs in the background — a single-task delegate_task from the top-level agent is now always dispatched async, so the parent turn returns immediately and the subagent's result re-enters the conversation when it finishes. - run_agent._dispatch_delegate_task (the live model path) forces background=True for top-level single-task calls; the schema-level `background` param is ignored. - A batch (tasks with >1 item) stays synchronous (fan-out can't go async). - A delegation from an orchestrator subagent (depth > 0) stays synchronous — it needs its workers' results within its own turn. - The function-level default is unchanged, so direct Python callers/tests keep the historical synchronous behavior. - On async-pool capacity rejection, single-task now falls through to a synchronous run instead of erroring (the child stays attached for interrupt propagation; detach happens only on a successful dispatch). - Schema `background` param marked deprecated/ignored; tool description updated to state the always-background single-task rule.
Extend the always-background behavior to the full fan-out. A batch is now dispatched as N independent async subagents (one handle each), instead of running synchronously. Single task and batch both return immediately; each subagent's result re-enters the conversation as its own message when it finishes. - delegate_task: when background is set, loop over ALL built children and dispatch each via dispatch_async_delegation; return a combined handle block (count + per-task delegation_ids). Children the async pool rejects (at capacity) run synchronously inline and are reported alongside the dispatched handles, so nothing is silently dropped. - run_agent._dispatch_delegate_task + registry handler: force background for any top-level model delegation (single OR batch); orchestrator subagents (depth > 0) still run synchronously since they need workers' results within their own turn. - Removed the v1 'batch async not supported' rejection. - Tool description updated: BOTH MODES RUN IN THE BACKGROUND. - Tests updated to assert batch fan-out dispatches each task async (verified E2E: 3-task batch -> 3 independent completion-queue events).
…d block
Correct the fan-out semantics: a backgrounded batch is dispatched as ONE
async unit (one handle, one async-pool slot), not N independent dispatches.
The unit runs all children in parallel, waits on every one, and emits a
SINGLE completion event carrying the consolidated per-task results. The chat
is never blocked; when all subagents finish, their full summaries re-enter
the conversation together as one message.
- async_delegation.dispatch_async_delegation_batch + _finalize_batch: a batch
occupies one slot; its runner returns the combined {results:[...]} dict and
one event with the full results list is pushed to the completion queue.
- delegate_tool: extract the sync execution+aggregation into
_execute_and_aggregate(); background dispatches it via the batch unit and
returns one handle; on pool-capacity rejection it runs the batch inline.
- process_registry._format_async_delegation: render a consolidated multi-task
block (TASK i/N + per-task summary) when the event carries is_batch/results.
- Tests updated; E2E verified: 3-task batch -> immediate return -> one combined
completion block with all three summaries.
teknium1
force-pushed
the
hermes/hermes-572aa955
branch
from
June 20, 2026 18:05
b897cce to
576b4b5
Compare
teknium1
changed the title
kpadilha
pushed a commit
to kpadilha/hermes-agent
that referenced
this pull request
Jun 24, 2026
…idated return (NousResearch#49734) * feat(delegation): single-task delegate_task always runs in the background The model no longer decides whether a subagent runs in the background — a single-task delegate_task from the top-level agent is now always dispatched async, so the parent turn returns immediately and the subagent's result re-enters the conversation when it finishes. - run_agent._dispatch_delegate_task (the live model path) forces background=True for top-level single-task calls; the schema-level `background` param is ignored. - A batch (tasks with >1 item) stays synchronous (fan-out can't go async). - A delegation from an orchestrator subagent (depth > 0) stays synchronous — it needs its workers' results within its own turn. - The function-level default is unchanged, so direct Python callers/tests keep the historical synchronous behavior. - On async-pool capacity rejection, single-task now falls through to a synchronous run instead of erroring (the child stays attached for interrupt propagation; detach happens only on a successful dispatch). - Schema `background` param marked deprecated/ignored; tool description updated to state the always-background single-task rule. * feat(delegation): all delegate_task fan-out runs in the background Extend the always-background behavior to the full fan-out. A batch is now dispatched as N independent async subagents (one handle each), instead of running synchronously. Single task and batch both return immediately; each subagent's result re-enters the conversation as its own message when it finishes. - delegate_task: when background is set, loop over ALL built children and dispatch each via dispatch_async_delegation; return a combined handle block (count + per-task delegation_ids). Children the async pool rejects (at capacity) run synchronously inline and are reported alongside the dispatched handles, so nothing is silently dropped. - run_agent._dispatch_delegate_task + registry handler: force background for any top-level model delegation (single OR batch); orchestrator subagents (depth > 0) still run synchronously since they need workers' results within their own turn. - Removed the v1 'batch async not supported' rejection. - Tool description updated: BOTH MODES RUN IN THE BACKGROUND. - Tests updated to assert batch fan-out dispatches each task async (verified E2E: 3-task batch -> 3 independent completion-queue events). * fix(delegation): background fan-out joins and returns one consolidated block Correct the fan-out semantics: a backgrounded batch is dispatched as ONE async unit (one handle, one async-pool slot), not N independent dispatches. The unit runs all children in parallel, waits on every one, and emits a SINGLE completion event carrying the consolidated per-task results. The chat is never blocked; when all subagents finish, their full summaries re-enter the conversation together as one message. - async_delegation.dispatch_async_delegation_batch + _finalize_batch: a batch occupies one slot; its runner returns the combined {results:[...]} dict and one event with the full results list is pushed to the completion queue. - delegate_tool: extract the sync execution+aggregation into _execute_and_aggregate(); background dispatches it via the batch unit and returns one handle; on pool-capacity rejection it runs the batch inline. - process_registry._format_async_delegation: render a consolidated multi-task block (TASK i/N + per-task summary) when the event carries is_batch/results. - Tests updated; E2E verified: 3-task batch -> immediate return -> one combined completion block with all three summaries.
pai-scaffolde
pushed a commit
to pai-scaffolde/hermes-agent
that referenced
this pull request
Jun 28, 2026
…idated return (NousResearch#49734) * feat(delegation): single-task delegate_task always runs in the background The model no longer decides whether a subagent runs in the background — a single-task delegate_task from the top-level agent is now always dispatched async, so the parent turn returns immediately and the subagent's result re-enters the conversation when it finishes. - run_agent._dispatch_delegate_task (the live model path) forces background=True for top-level single-task calls; the schema-level `background` param is ignored. - A batch (tasks with >1 item) stays synchronous (fan-out can't go async). - A delegation from an orchestrator subagent (depth > 0) stays synchronous — it needs its workers' results within its own turn. - The function-level default is unchanged, so direct Python callers/tests keep the historical synchronous behavior. - On async-pool capacity rejection, single-task now falls through to a synchronous run instead of erroring (the child stays attached for interrupt propagation; detach happens only on a successful dispatch). - Schema `background` param marked deprecated/ignored; tool description updated to state the always-background single-task rule. * feat(delegation): all delegate_task fan-out runs in the background Extend the always-background behavior to the full fan-out. A batch is now dispatched as N independent async subagents (one handle each), instead of running synchronously. Single task and batch both return immediately; each subagent's result re-enters the conversation as its own message when it finishes. - delegate_task: when background is set, loop over ALL built children and dispatch each via dispatch_async_delegation; return a combined handle block (count + per-task delegation_ids). Children the async pool rejects (at capacity) run synchronously inline and are reported alongside the dispatched handles, so nothing is silently dropped. - run_agent._dispatch_delegate_task + registry handler: force background for any top-level model delegation (single OR batch); orchestrator subagents (depth > 0) still run synchronously since they need workers' results within their own turn. - Removed the v1 'batch async not supported' rejection. - Tool description updated: BOTH MODES RUN IN THE BACKGROUND. - Tests updated to assert batch fan-out dispatches each task async (verified E2E: 3-task batch -> 3 independent completion-queue events). * fix(delegation): background fan-out joins and returns one consolidated block Correct the fan-out semantics: a backgrounded batch is dispatched as ONE async unit (one handle, one async-pool slot), not N independent dispatches. The unit runs all children in parallel, waits on every one, and emits a SINGLE completion event carrying the consolidated per-task results. The chat is never blocked; when all subagents finish, their full summaries re-enter the conversation together as one message. - async_delegation.dispatch_async_delegation_batch + _finalize_batch: a batch occupies one slot; its runner returns the combined {results:[...]} dict and one event with the full results list is pushed to the completion queue. - delegate_tool: extract the sync execution+aggregation into _execute_and_aggregate(); background dispatches it via the batch unit and returns one handle; on pool-capacity rejection it runs the batch inline. - process_registry._format_async_delegation: render a consolidated multi-task block (TASK i/N + per-task summary) when the event carries is_batch/results. - Tests updated; E2E verified: 3-task batch -> immediate return -> one combined completion block with all three summaries.
1 task
waefrebeorn
pushed a commit
to waefrebeorn/slermes
that referenced
this pull request
Jul 2, 2026
…idated return (NousResearch#49734) * feat(delegation): single-task delegate_task always runs in the background The model no longer decides whether a subagent runs in the background — a single-task delegate_task from the top-level agent is now always dispatched async, so the parent turn returns immediately and the subagent's result re-enters the conversation when it finishes. - run_agent._dispatch_delegate_task (the live model path) forces background=True for top-level single-task calls; the schema-level `background` param is ignored. - A batch (tasks with >1 item) stays synchronous (fan-out can't go async). - A delegation from an orchestrator subagent (depth > 0) stays synchronous — it needs its workers' results within its own turn. - The function-level default is unchanged, so direct Python callers/tests keep the historical synchronous behavior. - On async-pool capacity rejection, single-task now falls through to a synchronous run instead of erroring (the child stays attached for interrupt propagation; detach happens only on a successful dispatch). - Schema `background` param marked deprecated/ignored; tool description updated to state the always-background single-task rule. * feat(delegation): all delegate_task fan-out runs in the background Extend the always-background behavior to the full fan-out. A batch is now dispatched as N independent async subagents (one handle each), instead of running synchronously. Single task and batch both return immediately; each subagent's result re-enters the conversation as its own message when it finishes. - delegate_task: when background is set, loop over ALL built children and dispatch each via dispatch_async_delegation; return a combined handle block (count + per-task delegation_ids). Children the async pool rejects (at capacity) run synchronously inline and are reported alongside the dispatched handles, so nothing is silently dropped. - run_agent._dispatch_delegate_task + registry handler: force background for any top-level model delegation (single OR batch); orchestrator subagents (depth > 0) still run synchronously since they need workers' results within their own turn. - Removed the v1 'batch async not supported' rejection. - Tool description updated: BOTH MODES RUN IN THE BACKGROUND. - Tests updated to assert batch fan-out dispatches each task async (verified E2E: 3-task batch -> 3 independent completion-queue events). * fix(delegation): background fan-out joins and returns one consolidated block Correct the fan-out semantics: a backgrounded batch is dispatched as ONE async unit (one handle, one async-pool slot), not N independent dispatches. The unit runs all children in parallel, waits on every one, and emits a SINGLE completion event carrying the consolidated per-task results. The chat is never blocked; when all subagents finish, their full summaries re-enter the conversation together as one message. - async_delegation.dispatch_async_delegation_batch + _finalize_batch: a batch occupies one slot; its runner returns the combined {results:[...]} dict and one event with the full results list is pushed to the completion queue. - delegate_tool: extract the sync execution+aggregation into _execute_and_aggregate(); background dispatches it via the batch unit and returns one handle; on pool-capacity rejection it runs the batch inline. - process_registry._format_async_delegation: render a consolidated multi-task block (TASK i/N + per-task summary) when the event carries is_batch/results. - Tests updated; E2E verified: 3-task batch -> immediate return -> one combined completion block with all three summaries.
sliceintoit
pushed a commit
to sliceintoit/hermes-agent
that referenced
this pull request
Jul 4, 2026
…idated return (NousResearch#49734) * feat(delegation): single-task delegate_task always runs in the background The model no longer decides whether a subagent runs in the background — a single-task delegate_task from the top-level agent is now always dispatched async, so the parent turn returns immediately and the subagent's result re-enters the conversation when it finishes. - run_agent._dispatch_delegate_task (the live model path) forces background=True for top-level single-task calls; the schema-level `background` param is ignored. - A batch (tasks with >1 item) stays synchronous (fan-out can't go async). - A delegation from an orchestrator subagent (depth > 0) stays synchronous — it needs its workers' results within its own turn. - The function-level default is unchanged, so direct Python callers/tests keep the historical synchronous behavior. - On async-pool capacity rejection, single-task now falls through to a synchronous run instead of erroring (the child stays attached for interrupt propagation; detach happens only on a successful dispatch). - Schema `background` param marked deprecated/ignored; tool description updated to state the always-background single-task rule. * feat(delegation): all delegate_task fan-out runs in the background Extend the always-background behavior to the full fan-out. A batch is now dispatched as N independent async subagents (one handle each), instead of running synchronously. Single task and batch both return immediately; each subagent's result re-enters the conversation as its own message when it finishes. - delegate_task: when background is set, loop over ALL built children and dispatch each via dispatch_async_delegation; return a combined handle block (count + per-task delegation_ids). Children the async pool rejects (at capacity) run synchronously inline and are reported alongside the dispatched handles, so nothing is silently dropped. - run_agent._dispatch_delegate_task + registry handler: force background for any top-level model delegation (single OR batch); orchestrator subagents (depth > 0) still run synchronously since they need workers' results within their own turn. - Removed the v1 'batch async not supported' rejection. - Tool description updated: BOTH MODES RUN IN THE BACKGROUND. - Tests updated to assert batch fan-out dispatches each task async (verified E2E: 3-task batch -> 3 independent completion-queue events). * fix(delegation): background fan-out joins and returns one consolidated block Correct the fan-out semantics: a backgrounded batch is dispatched as ONE async unit (one handle, one async-pool slot), not N independent dispatches. The unit runs all children in parallel, waits on every one, and emits a SINGLE completion event carrying the consolidated per-task results. The chat is never blocked; when all subagents finish, their full summaries re-enter the conversation together as one message. - async_delegation.dispatch_async_delegation_batch + _finalize_batch: a batch occupies one slot; its runner returns the combined {results:[...]} dict and one event with the full results list is pushed to the completion queue. - delegate_tool: extract the sync execution+aggregation into _execute_and_aggregate(); background dispatches it via the batch unit and returns one handle; on pool-capacity rejection it runs the batch inline. - process_registry._format_async_delegation: render a consolidated multi-task block (TASK i/N + per-task summary) when the event carries is_batch/results. - Tests updated; E2E verified: 3-task batch -> immediate return -> one combined completion block with all three summaries.
habarmc1223-sudo
pushed a commit
to habarmc1223-sudo/hermes-agent-fluxmem
that referenced
this pull request
Jul 8, 2026
…idated return (NousResearch#49734) * feat(delegation): single-task delegate_task always runs in the background The model no longer decides whether a subagent runs in the background — a single-task delegate_task from the top-level agent is now always dispatched async, so the parent turn returns immediately and the subagent's result re-enters the conversation when it finishes. - run_agent._dispatch_delegate_task (the live model path) forces background=True for top-level single-task calls; the schema-level `background` param is ignored. - A batch (tasks with >1 item) stays synchronous (fan-out can't go async). - A delegation from an orchestrator subagent (depth > 0) stays synchronous — it needs its workers' results within its own turn. - The function-level default is unchanged, so direct Python callers/tests keep the historical synchronous behavior. - On async-pool capacity rejection, single-task now falls through to a synchronous run instead of erroring (the child stays attached for interrupt propagation; detach happens only on a successful dispatch). - Schema `background` param marked deprecated/ignored; tool description updated to state the always-background single-task rule. * feat(delegation): all delegate_task fan-out runs in the background Extend the always-background behavior to the full fan-out. A batch is now dispatched as N independent async subagents (one handle each), instead of running synchronously. Single task and batch both return immediately; each subagent's result re-enters the conversation as its own message when it finishes. - delegate_task: when background is set, loop over ALL built children and dispatch each via dispatch_async_delegation; return a combined handle block (count + per-task delegation_ids). Children the async pool rejects (at capacity) run synchronously inline and are reported alongside the dispatched handles, so nothing is silently dropped. - run_agent._dispatch_delegate_task + registry handler: force background for any top-level model delegation (single OR batch); orchestrator subagents (depth > 0) still run synchronously since they need workers' results within their own turn. - Removed the v1 'batch async not supported' rejection. - Tool description updated: BOTH MODES RUN IN THE BACKGROUND. - Tests updated to assert batch fan-out dispatches each task async (verified E2E: 3-task batch -> 3 independent completion-queue events). * fix(delegation): background fan-out joins and returns one consolidated block Correct the fan-out semantics: a backgrounded batch is dispatched as ONE async unit (one handle, one async-pool slot), not N independent dispatches. The unit runs all children in parallel, waits on every one, and emits a SINGLE completion event carrying the consolidated per-task results. The chat is never blocked; when all subagents finish, their full summaries re-enter the conversation together as one message. - async_delegation.dispatch_async_delegation_batch + _finalize_batch: a batch occupies one slot; its runner returns the combined {results:[...]} dict and one event with the full results list is pushed to the completion queue. - delegate_tool: extract the sync execution+aggregation into _execute_and_aggregate(); background dispatches it via the batch unit and returns one handle; on pool-capacity rejection it runs the batch inline. - process_registry._format_async_delegation: render a consolidated multi-task block (TASK i/N + per-task summary) when the event carries is_batch/results. - Tests updated; E2E verified: 3-task batch -> immediate return -> one combined completion block with all three summaries.
santhreal
pushed a commit
to santhreal/hermes-agent
that referenced
this pull request
Jul 13, 2026
…idated return (NousResearch#49734) * feat(delegation): single-task delegate_task always runs in the background The model no longer decides whether a subagent runs in the background — a single-task delegate_task from the top-level agent is now always dispatched async, so the parent turn returns immediately and the subagent's result re-enters the conversation when it finishes. - run_agent._dispatch_delegate_task (the live model path) forces background=True for top-level single-task calls; the schema-level `background` param is ignored. - A batch (tasks with >1 item) stays synchronous (fan-out can't go async). - A delegation from an orchestrator subagent (depth > 0) stays synchronous — it needs its workers' results within its own turn. - The function-level default is unchanged, so direct Python callers/tests keep the historical synchronous behavior. - On async-pool capacity rejection, single-task now falls through to a synchronous run instead of erroring (the child stays attached for interrupt propagation; detach happens only on a successful dispatch). - Schema `background` param marked deprecated/ignored; tool description updated to state the always-background single-task rule. * feat(delegation): all delegate_task fan-out runs in the background Extend the always-background behavior to the full fan-out. A batch is now dispatched as N independent async subagents (one handle each), instead of running synchronously. Single task and batch both return immediately; each subagent's result re-enters the conversation as its own message when it finishes. - delegate_task: when background is set, loop over ALL built children and dispatch each via dispatch_async_delegation; return a combined handle block (count + per-task delegation_ids). Children the async pool rejects (at capacity) run synchronously inline and are reported alongside the dispatched handles, so nothing is silently dropped. - run_agent._dispatch_delegate_task + registry handler: force background for any top-level model delegation (single OR batch); orchestrator subagents (depth > 0) still run synchronously since they need workers' results within their own turn. - Removed the v1 'batch async not supported' rejection. - Tool description updated: BOTH MODES RUN IN THE BACKGROUND. - Tests updated to assert batch fan-out dispatches each task async (verified E2E: 3-task batch -> 3 independent completion-queue events). * fix(delegation): background fan-out joins and returns one consolidated block Correct the fan-out semantics: a backgrounded batch is dispatched as ONE async unit (one handle, one async-pool slot), not N independent dispatches. The unit runs all children in parallel, waits on every one, and emits a SINGLE completion event carrying the consolidated per-task results. The chat is never blocked; when all subagents finish, their full summaries re-enter the conversation together as one message. - async_delegation.dispatch_async_delegation_batch + _finalize_batch: a batch occupies one slot; its runner returns the combined {results:[...]} dict and one event with the full results list is pushed to the completion queue. - delegate_tool: extract the sync execution+aggregation into _execute_and_aggregate(); background dispatches it via the batch unit and returns one handle; on pool-capacity rejection it runs the batch inline. - process_registry._format_async_delegation: render a consolidated multi-task block (TASK i/N + per-task summary) when the event carries is_batch/results. - Tests updated; E2E verified: 3-task batch -> immediate return -> one combined completion block with all three summaries.
This was referenced Jul 14, 2026
19 tasks
Gravezzz
pushed a commit
to Gravezzz/hermes-agent
that referenced
this pull request
Jul 21, 2026
…idated return (NousResearch#49734) * feat(delegation): single-task delegate_task always runs in the background The model no longer decides whether a subagent runs in the background — a single-task delegate_task from the top-level agent is now always dispatched async, so the parent turn returns immediately and the subagent's result re-enters the conversation when it finishes. - run_agent._dispatch_delegate_task (the live model path) forces background=True for top-level single-task calls; the schema-level `background` param is ignored. - A batch (tasks with >1 item) stays synchronous (fan-out can't go async). - A delegation from an orchestrator subagent (depth > 0) stays synchronous — it needs its workers' results within its own turn. - The function-level default is unchanged, so direct Python callers/tests keep the historical synchronous behavior. - On async-pool capacity rejection, single-task now falls through to a synchronous run instead of erroring (the child stays attached for interrupt propagation; detach happens only on a successful dispatch). - Schema `background` param marked deprecated/ignored; tool description updated to state the always-background single-task rule. * feat(delegation): all delegate_task fan-out runs in the background Extend the always-background behavior to the full fan-out. A batch is now dispatched as N independent async subagents (one handle each), instead of running synchronously. Single task and batch both return immediately; each subagent's result re-enters the conversation as its own message when it finishes. - delegate_task: when background is set, loop over ALL built children and dispatch each via dispatch_async_delegation; return a combined handle block (count + per-task delegation_ids). Children the async pool rejects (at capacity) run synchronously inline and are reported alongside the dispatched handles, so nothing is silently dropped. - run_agent._dispatch_delegate_task + registry handler: force background for any top-level model delegation (single OR batch); orchestrator subagents (depth > 0) still run synchronously since they need workers' results within their own turn. - Removed the v1 'batch async not supported' rejection. - Tool description updated: BOTH MODES RUN IN THE BACKGROUND. - Tests updated to assert batch fan-out dispatches each task async (verified E2E: 3-task batch -> 3 independent completion-queue events). * fix(delegation): background fan-out joins and returns one consolidated block Correct the fan-out semantics: a backgrounded batch is dispatched as ONE async unit (one handle, one async-pool slot), not N independent dispatches. The unit runs all children in parallel, waits on every one, and emits a SINGLE completion event carrying the consolidated per-task results. The chat is never blocked; when all subagents finish, their full summaries re-enter the conversation together as one message. - async_delegation.dispatch_async_delegation_batch + _finalize_batch: a batch occupies one slot; its runner returns the combined {results:[...]} dict and one event with the full results list is pushed to the completion queue. - delegate_tool: extract the sync execution+aggregation into _execute_and_aggregate(); background dispatches it via the batch unit and returns one handle; on pool-capacity rejection it runs the batch inline. - process_registry._format_async_delegation: render a consolidated multi-task block (TASK i/N + per-task summary) when the event carries is_batch/results. - Tests updated; E2E verified: 3-task batch -> immediate return -> one combined completion block with all three summaries.
leewenjie
pushed a commit
to leewenjie/hermes-agent
that referenced
this pull request
Aug 7, 2026
…idated return (NousResearch#49734) * feat(delegation): single-task delegate_task always runs in the background The model no longer decides whether a subagent runs in the background — a single-task delegate_task from the top-level agent is now always dispatched async, so the parent turn returns immediately and the subagent's result re-enters the conversation when it finishes. - run_agent._dispatch_delegate_task (the live model path) forces background=True for top-level single-task calls; the schema-level `background` param is ignored. - A batch (tasks with >1 item) stays synchronous (fan-out can't go async). - A delegation from an orchestrator subagent (depth > 0) stays synchronous — it needs its workers' results within its own turn. - The function-level default is unchanged, so direct Python callers/tests keep the historical synchronous behavior. - On async-pool capacity rejection, single-task now falls through to a synchronous run instead of erroring (the child stays attached for interrupt propagation; detach happens only on a successful dispatch). - Schema `background` param marked deprecated/ignored; tool description updated to state the always-background single-task rule. * feat(delegation): all delegate_task fan-out runs in the background Extend the always-background behavior to the full fan-out. A batch is now dispatched as N independent async subagents (one handle each), instead of running synchronously. Single task and batch both return immediately; each subagent's result re-enters the conversation as its own message when it finishes. - delegate_task: when background is set, loop over ALL built children and dispatch each via dispatch_async_delegation; return a combined handle block (count + per-task delegation_ids). Children the async pool rejects (at capacity) run synchronously inline and are reported alongside the dispatched handles, so nothing is silently dropped. - run_agent._dispatch_delegate_task + registry handler: force background for any top-level model delegation (single OR batch); orchestrator subagents (depth > 0) still run synchronously since they need workers' results within their own turn. - Removed the v1 'batch async not supported' rejection. - Tool description updated: BOTH MODES RUN IN THE BACKGROUND. - Tests updated to assert batch fan-out dispatches each task async (verified E2E: 3-task batch -> 3 independent completion-queue events). * fix(delegation): background fan-out joins and returns one consolidated block Correct the fan-out semantics: a backgrounded batch is dispatched as ONE async unit (one handle, one async-pool slot), not N independent dispatches. The unit runs all children in parallel, waits on every one, and emits a SINGLE completion event carrying the consolidated per-task results. The chat is never blocked; when all subagents finish, their full summaries re-enter the conversation together as one message. - async_delegation.dispatch_async_delegation_batch + _finalize_batch: a batch occupies one slot; its runner returns the combined {results:[...]} dict and one event with the full results list is pushed to the completion queue. - delegate_tool: extract the sync execution+aggregation into _execute_and_aggregate(); background dispatches it via the batch unit and returns one handle; on pool-capacity rejection it runs the batch inline. - process_registry._format_async_delegation: render a consolidated multi-task block (TASK i/N + per-task summary) when the event carries is_batch/results. - Tests updated; E2E verified: 3-task batch -> immediate return -> one combined completion block with all three summaries.
1 task
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 join this conversation on GitHub.
Already have an account?
Sign in to comment
2 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.
Summary
A
delegate_taskfan-out now runs in the background — the chat is never blocked — and when all subagents finish, their summaries come back together as one consolidated block. The main agent calls delegate, N subagents spawn and run in parallel, the user keeps chatting, and once every subagent completes the combined results re-enter the conversation as a single message.The model no longer decides whether to background; it's automatic for any top-level delegation.
How it works
Changes
tools/async_delegation.py: newdispatch_async_delegation_batch+_finalize_batch. The whole fan-out occupies one async-pool slot and one handle; its runner runs/join all children and returns the combined{results:[...]}; a single completion event carrying the full per-task results list is pushed to the shared completion queue.tools/delegate_tool.py: extracted the existing sync execution + aggregation (thread-pool parallel run, join, memory/hook/cost rollup) into_execute_and_aggregate(). The background path dispatches that whole function as one batch unit and returns one handle; the synchronous path calls it directly. On async-pool capacity rejection, the batch runs inline.run_agent._dispatch_delegate_task+ registry handler: force background for any top-level model delegation (single or batch). Orchestrator subagents (_delegate_depth > 0) stay synchronous — they need their workers' results within their own turn.tools/process_registry._format_async_delegation: renders a consolidated multi-task block (TASK i/N+ per-task summary) when the completion event carriesis_batch/results.delegate_task()keep the synchronous default; only the model-dispatch boundary forces background.Validation
Targeted tests:
test_async_delegation.py,test_delegate.py,test_process_registry.py— 230 passed, 0 failed. E2E (real dispatch + completion queue): a 3-task batch returns a handle immediately and emits exactly one combined completion event with all three summaries rendered together.Infographic