Feat nudge improvement #892
Open
+411
−4
Conversation
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
Document the approved design for fixing subordinate agent nudge behavior and implementing auto-nudge detection. Co-Authored-By: Claude Opus 4.5 <[email protected]>
Problem: Nudging a subordinate agent incorrectly reset execution to Agent 0 because streaming_agent was cleared at monologue end. Solution: - Add last_active_agent field that persists beyond monologue completion - Update get_agent() to check last_active_agent before falling back - Nudge now resumes from the last active agent, not agent0 Auto-nudge feature: - Add configurable watchdog that monitors LLM streaming timeouts - If no streaming chunks for auto_nudge_timeout seconds, trigger nudge - Disabled by default (auto_nudge_enabled: false) API improvement: - Nudge response now includes which agent was nudged Co-Authored-By: Claude Opus 4.5 <[email protected]>
Tests verify: - get_agent() priority: streaming_agent > last_active_agent > agent0 - nudge() restarts correct agent (subordinate, not always agent0) - auto_nudge config fields exist with correct defaults 10 tests covering the core nudge fix and auto-nudge config. Co-Authored-By: Claude Opus 4.5 <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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Outdated suggestions cannot be applied.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
Problem
When a subordinate agent (Agent 1, 2, etc.) was running and the user pressed Nudge, the entire execution reset to Agent 0. This happened because streaming_agent was cleared at the end of each monologue, causing get_agent() to fall back to agent0.
Solution
Added last_active_agent field that persists beyond monologue completion:
def get_agent(self):
return self.streaming_agent or self.last_active_agent or self.agent0
Changes
Auto-Nudge Feature
Configurable watchdog that monitors LLM streaming timeouts:
auto_nudge_enabled: bool = False # off by default
auto_nudge_timeout: int = 60 # seconds
Test plan