feat(moa): show each reference model's output as a labelled block before the aggregator - #53793
Merged
Conversation
…ore the aggregator When a MoA preset is selected, each reference model's answer now renders in the CLI as a thinking-style block labelled with its source model, BEFORE the aggregator responds — so the mixture-of-agents process is visible instead of a silent pause. The aggregator's response (and its tool actions) follow as normal. Mechanism (shared seam, all surfaces): - MoAChatCompletions/MoAClient take an optional reference_callback and emit 'moa.reference' (index/count/label/text) per reference, then 'moa.aggregating' (aggregator label) once. agent_init wires this to the agent's tool_progress_callback, which every surface already consumes — so the events reach CLI/TUI/desktop/gateway with no new plumbing. - CLI _on_tool_progress renders 'moa.reference' as a labelled '┊ ◇ Reference i/n — <model>' header + a thinking-style preview (reusing _emit_reasoning_ preview), and 'moa.aggregating' as a spinner transition. Display-only; never touches message history (cache-safe). Turn-scoped reference cache: the agent loop calls the facade once per tool-loop iteration, but the advisory message view is identical across iterations within a turn, so references are now run AND displayed once per user turn (keyed by the advisory view's signature) instead of re-running/re-spamming on every iteration. This also cuts reference API cost from O(iterations) back to O(turns). Verified live via interactive PTY on the opus-gpt preset (gpt-5.5 + opus refs): reference blocks render once per turn, labelled by model, before the aggregator; fresh blocks on each new turn; aggregator tool actions still execute. Follow-up: TUI/desktop rich rendering + gateway batched-summary already receive the events via tool_progress_callback; their surface-specific renderers are a separate change.
teknium1
force-pushed
the
hermes/moa-reference-blocks
branch
from
June 27, 2026 19:30
386c30a to
0ba4c5f
Compare
alt-glitch
added
type/feature
pai-scaffolde
pushed a commit
to pai-scaffolde/hermes-agent
that referenced
this pull request
Jun 28, 2026
…ore the aggregator (NousResearch#53793) When a MoA preset is selected, each reference model's answer now renders in the CLI as a thinking-style block labelled with its source model, BEFORE the aggregator responds — so the mixture-of-agents process is visible instead of a silent pause. The aggregator's response (and its tool actions) follow as normal. Mechanism (shared seam, all surfaces): - MoAChatCompletions/MoAClient take an optional reference_callback and emit 'moa.reference' (index/count/label/text) per reference, then 'moa.aggregating' (aggregator label) once. agent_init wires this to the agent's tool_progress_callback, which every surface already consumes — so the events reach CLI/TUI/desktop/gateway with no new plumbing. - CLI _on_tool_progress renders 'moa.reference' as a labelled '┊ ◇ Reference i/n — <model>' header + a thinking-style preview (reusing _emit_reasoning_ preview), and 'moa.aggregating' as a spinner transition. Display-only; never touches message history (cache-safe). Turn-scoped reference cache: the agent loop calls the facade once per tool-loop iteration, but the advisory message view is identical across iterations within a turn, so references are now run AND displayed once per user turn (keyed by the advisory view's signature) instead of re-running/re-spamming on every iteration. This also cuts reference API cost from O(iterations) back to O(turns). Verified live via interactive PTY on the opus-gpt preset (gpt-5.5 + opus refs): reference blocks render once per turn, labelled by model, before the aggregator; fresh blocks on each new turn; aggregator tool actions still execute. Follow-up: TUI/desktop rich rendering + gateway batched-summary already receive the events via tool_progress_callback; their surface-specific renderers are a separate change.
1 task
waefrebeorn
pushed a commit
to waefrebeorn/slermes
that referenced
this pull request
Jul 2, 2026
…ore the aggregator (NousResearch#53793) When a MoA preset is selected, each reference model's answer now renders in the CLI as a thinking-style block labelled with its source model, BEFORE the aggregator responds — so the mixture-of-agents process is visible instead of a silent pause. The aggregator's response (and its tool actions) follow as normal. Mechanism (shared seam, all surfaces): - MoAChatCompletions/MoAClient take an optional reference_callback and emit 'moa.reference' (index/count/label/text) per reference, then 'moa.aggregating' (aggregator label) once. agent_init wires this to the agent's tool_progress_callback, which every surface already consumes — so the events reach CLI/TUI/desktop/gateway with no new plumbing. - CLI _on_tool_progress renders 'moa.reference' as a labelled '┊ ◇ Reference i/n — <model>' header + a thinking-style preview (reusing _emit_reasoning_ preview), and 'moa.aggregating' as a spinner transition. Display-only; never touches message history (cache-safe). Turn-scoped reference cache: the agent loop calls the facade once per tool-loop iteration, but the advisory message view is identical across iterations within a turn, so references are now run AND displayed once per user turn (keyed by the advisory view's signature) instead of re-running/re-spamming on every iteration. This also cuts reference API cost from O(iterations) back to O(turns). Verified live via interactive PTY on the opus-gpt preset (gpt-5.5 + opus refs): reference blocks render once per turn, labelled by model, before the aggregator; fresh blocks on each new turn; aggregator tool actions still execute. Follow-up: TUI/desktop rich rendering + gateway batched-summary already receive the events via tool_progress_callback; their surface-specific renderers are a separate change.
habarmc1223-sudo
pushed a commit
to habarmc1223-sudo/hermes-agent-fluxmem
that referenced
this pull request
Jul 8, 2026
…ore the aggregator (NousResearch#53793) When a MoA preset is selected, each reference model's answer now renders in the CLI as a thinking-style block labelled with its source model, BEFORE the aggregator responds — so the mixture-of-agents process is visible instead of a silent pause. The aggregator's response (and its tool actions) follow as normal. Mechanism (shared seam, all surfaces): - MoAChatCompletions/MoAClient take an optional reference_callback and emit 'moa.reference' (index/count/label/text) per reference, then 'moa.aggregating' (aggregator label) once. agent_init wires this to the agent's tool_progress_callback, which every surface already consumes — so the events reach CLI/TUI/desktop/gateway with no new plumbing. - CLI _on_tool_progress renders 'moa.reference' as a labelled '┊ ◇ Reference i/n — <model>' header + a thinking-style preview (reusing _emit_reasoning_ preview), and 'moa.aggregating' as a spinner transition. Display-only; never touches message history (cache-safe). Turn-scoped reference cache: the agent loop calls the facade once per tool-loop iteration, but the advisory message view is identical across iterations within a turn, so references are now run AND displayed once per user turn (keyed by the advisory view's signature) instead of re-running/re-spamming on every iteration. This also cuts reference API cost from O(iterations) back to O(turns). Verified live via interactive PTY on the opus-gpt preset (gpt-5.5 + opus refs): reference blocks render once per turn, labelled by model, before the aggregator; fresh blocks on each new turn; aggregator tool actions still execute. Follow-up: TUI/desktop rich rendering + gateway batched-summary already receive the events via tool_progress_callback; their surface-specific renderers are a separate change.
santhreal
pushed a commit
to santhreal/hermes-agent
that referenced
this pull request
Jul 13, 2026
…ore the aggregator (NousResearch#53793) When a MoA preset is selected, each reference model's answer now renders in the CLI as a thinking-style block labelled with its source model, BEFORE the aggregator responds — so the mixture-of-agents process is visible instead of a silent pause. The aggregator's response (and its tool actions) follow as normal. Mechanism (shared seam, all surfaces): - MoAChatCompletions/MoAClient take an optional reference_callback and emit 'moa.reference' (index/count/label/text) per reference, then 'moa.aggregating' (aggregator label) once. agent_init wires this to the agent's tool_progress_callback, which every surface already consumes — so the events reach CLI/TUI/desktop/gateway with no new plumbing. - CLI _on_tool_progress renders 'moa.reference' as a labelled '┊ ◇ Reference i/n — <model>' header + a thinking-style preview (reusing _emit_reasoning_ preview), and 'moa.aggregating' as a spinner transition. Display-only; never touches message history (cache-safe). Turn-scoped reference cache: the agent loop calls the facade once per tool-loop iteration, but the advisory message view is identical across iterations within a turn, so references are now run AND displayed once per user turn (keyed by the advisory view's signature) instead of re-running/re-spamming on every iteration. This also cuts reference API cost from O(iterations) back to O(turns). Verified live via interactive PTY on the opus-gpt preset (gpt-5.5 + opus refs): reference blocks render once per turn, labelled by model, before the aggregator; fresh blocks on each new turn; aggregator tool actions still execute. Follow-up: TUI/desktop rich rendering + gateway batched-summary already receive the events via tool_progress_callback; their surface-specific renderers are a separate change.
Gravezzz
pushed a commit
to Gravezzz/hermes-agent
that referenced
this pull request
Jul 21, 2026
…ore the aggregator (NousResearch#53793) When a MoA preset is selected, each reference model's answer now renders in the CLI as a thinking-style block labelled with its source model, BEFORE the aggregator responds — so the mixture-of-agents process is visible instead of a silent pause. The aggregator's response (and its tool actions) follow as normal. Mechanism (shared seam, all surfaces): - MoAChatCompletions/MoAClient take an optional reference_callback and emit 'moa.reference' (index/count/label/text) per reference, then 'moa.aggregating' (aggregator label) once. agent_init wires this to the agent's tool_progress_callback, which every surface already consumes — so the events reach CLI/TUI/desktop/gateway with no new plumbing. - CLI _on_tool_progress renders 'moa.reference' as a labelled '┊ ◇ Reference i/n — <model>' header + a thinking-style preview (reusing _emit_reasoning_ preview), and 'moa.aggregating' as a spinner transition. Display-only; never touches message history (cache-safe). Turn-scoped reference cache: the agent loop calls the facade once per tool-loop iteration, but the advisory message view is identical across iterations within a turn, so references are now run AND displayed once per user turn (keyed by the advisory view's signature) instead of re-running/re-spamming on every iteration. This also cuts reference API cost from O(iterations) back to O(turns). Verified live via interactive PTY on the opus-gpt preset (gpt-5.5 + opus refs): reference blocks render once per turn, labelled by model, before the aggregator; fresh blocks on each new turn; aggregator tool actions still execute. Follow-up: TUI/desktop rich rendering + gateway batched-summary already receive the events via tool_progress_callback; their surface-specific renderers are a separate change.
12 tasks
leewenjie
pushed a commit
to leewenjie/hermes-agent
that referenced
this pull request
Aug 7, 2026
…ore the aggregator (NousResearch#53793) When a MoA preset is selected, each reference model's answer now renders in the CLI as a thinking-style block labelled with its source model, BEFORE the aggregator responds — so the mixture-of-agents process is visible instead of a silent pause. The aggregator's response (and its tool actions) follow as normal. Mechanism (shared seam, all surfaces): - MoAChatCompletions/MoAClient take an optional reference_callback and emit 'moa.reference' (index/count/label/text) per reference, then 'moa.aggregating' (aggregator label) once. agent_init wires this to the agent's tool_progress_callback, which every surface already consumes — so the events reach CLI/TUI/desktop/gateway with no new plumbing. - CLI _on_tool_progress renders 'moa.reference' as a labelled '┊ ◇ Reference i/n — <model>' header + a thinking-style preview (reusing _emit_reasoning_ preview), and 'moa.aggregating' as a spinner transition. Display-only; never touches message history (cache-safe). Turn-scoped reference cache: the agent loop calls the facade once per tool-loop iteration, but the advisory message view is identical across iterations within a turn, so references are now run AND displayed once per user turn (keyed by the advisory view's signature) instead of re-running/re-spamming on every iteration. This also cuts reference API cost from O(iterations) back to O(turns). Verified live via interactive PTY on the opus-gpt preset (gpt-5.5 + opus refs): reference blocks render once per turn, labelled by model, before the aggregator; fresh blocks on each new turn; aggregator tool actions still execute. Follow-up: TUI/desktop rich rendering + gateway batched-summary already receive the events via tool_progress_callback; their surface-specific renderers are a separate change.
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
When a MoA preset is selected, each reference model's output now renders in the CLI as a thinking-style block labelled with its source model, before the aggregator responds — so the mixture-of-agents process is visible instead of a silent pause. The aggregator's response (and its tool actions) follow as normal.
Changes
agent/moa_loop.py:MoAChatCompletions/MoAClienttake an optionalreference_callbackand emitmoa.reference(index / count / label / text) per reference, then onemoa.aggregating(aggregator label). Added a turn-scoped reference cache: the agent loop calls the facade once per tool-loop iteration, but the advisory message view is identical across iterations within a turn, so references now run and display once per user turn (keyed by the advisory view's signature) instead of re-running and re-spamming on every iteration — also cutting reference API cost from O(iterations) to O(turns).agent/agent_init.py: wire the facade'sreference_callbackto the agent'stool_progress_callback(the seam every surface already consumes), so the events reach CLI/TUI/desktop/gateway with no new plumbing.cli.py:_on_tool_progressrendersmoa.referenceas a labelled┊ ◇ Reference i/n — <model>header + a thinking-style preview (reusing_emit_reasoning_preview), andmoa.aggregatingas a spinner transition. Display-only — never touches message history (cache-safe).tests/run_agent/test_moa_loop_mode.py); CLI rendering of reference/aggregating events (tests/cli/test_tool_progress_scrollback.py).Validation
Live interactive PTY on the
opus-gptpreset (gpt-5.5 + opus references):Tests:
tests/run_agent/test_moa_loop_mode.py+tests/cli/test_tool_progress_scrollback.py→ 29/29 pass.Scope / follow-up
This lands the CLI surface (verified live) plus the shared seam. TUI/desktop rich rendering and the gateway batched-summary already receive the
moa.reference/moa.aggregatingevents viatool_progress_callback; their surface-specific renderers are a focused follow-up.Infographic