stop-hook-fires-on-every-assistant-turn-not-just-stop

The session-close stop hook was triggering repeatedly on every assistant message during an active waiting state, not only when the agent actually initiated a stop() call. This produced 30+ false-positive evaluations flooding the conversation with hook feedback. The hook condition ‘Before allowing the agent to stop’ is being evaluated on idle assistant turns rather than gated on an actual stop event.

Related

• stop-hook-fires-on-every-turn-not-just-agent-stop
• stop-hook-fires-false-positives-on-agent-waiting-state
• stop-hook-false-positive-during-active-wait-state
• stop-hook-false-positive-loop-on-waiting-state
• stop-hook-fires-on-agent-idle-waiting-state-false-positive
• stop-hook-fires-on-waiting-state-not-just-on-stop
• stop-hook-fires-on-every-assistant-turn-not-just-stop-calls