stop-hook-fires-on-user-wait-not-just-agent-stop

The session-stop hook fires repeatedly whenever the agent emits a short reply and awaits user input (e.g., a one-sentence clarifying question), not only when the agent actually terminates. This produced ~30 false-positive evaluations in a single brainstorming session, flooding context with irrelevant hook feedback. Root cause: the hook runtime interprets the agent going quiet as a stop event, regardless of whether a stop() call was made. Fix: the hook script must detect whether there is an outstanding tool-use or user-input-await state before evaluating completion gates.

Related

• stop-hook-fires-false-positives-on-agent-waiting-state
• stop-hook-fires-on-agent-idle-waiting-state-false-positive
• stop-hook-false-positive-during-agent-wait-states
• stop-hook-false-positive-during-waiting-state
• stop-hook-fires-on-idle-wait-state-false-positive-loop
• stop-hook-fires-repeatedly-on-passive-agent-waiting-state
• stop-hook-fires-on-waiting-state-not-just-on-stop
• ucg-must-be-explicitly-invoked-not-implied