{"version":"2.1.292","anchor":"shell-task-kill-on-agent-exit-refactored-memory-pressure-lo","canonical_anchor":"shell-task-kill-on-agent-exit-refactored-memory-pressure-lo","heading":"Clearer log line when a finishing agent's shell tasks are stopped","tier":"internal","area":"Background Tasks","scope":"individual","heads_up":false,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.292\/e\/shell-task-kill-on-agent-exit-refactored-memory-pressure-lo","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.292","markdown":"### Clearer log line when a finishing agent's shell tasks are stopped\n\nThe log line for stopping an exiting agent's leftover shell tasks now names the task, and one extra cleanup step was removed\n\n**What**\n\nWhen an agent finishes, Claude Code stops any shell tasks (commands running in the background) that the agent started and left running. The log message it writes for each one now shows the task's id. Before, it printed the whole internal list of tasks instead.\n\nA separate extra cleanup step that ran after this loop has been removed.\n\nAlso, a log message about skipping cleanup under memory pressure is now limited per skip reason. Before, the limit was tied to a value that was never set.\n\n**Why**\n\nMost of this makes logs easier to read. The removed cleanup step is the one part that could change behaviour, so leftover shell tasks after an agent exits are worth watching for.\n\n- Area: Background Tasks\n- Tier: Under the hood\n- Useful: 1\/5\n- Signal: 0\/5\n- Scope: individual\n- Heads-up: no"}