{"version":"2.1.317","anchor":"repeat-command-detection-now-works-across-agents","canonical_anchor":"repeat-command-detection-now-works-across-agents","heading":"Claude Code now notices when one agent repeats another agent's command","tier":"internal","area":"Subagents","scope":"individual","heads_up":false,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.317\/e\/repeat-command-detection-now-works-across-agents","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.317","markdown":"### Claude Code now notices when one agent repeats another agent's command\n\nClaude Code's tracking of repeated commands now records which agent ran each one, so repeats across the main agent and subagents are noted\n\n**Unclear.** It is unclear whether Claude Code shows anything to you when it detects such a repeat.\n\n**What**\n\nClaude Code keeps track of commands that get run more than once. It now remembers which agent ran each command, so it can notice when a different agent repeats it. An agent here is either the main Claude Code conversation or a subagent, a helper it starts for part of a task. For example, the main agent running a command a subagent already ran now counts as a repeat.\n\nBefore, only repeats within the same agent's history were compared. These repeats are recorded in Claude Code's usage data, along with whether a subagent was working in its own separate copy of the project.\n\n**Why**\n\nThis lays the ground for keeping agents from redoing each other's work.\n\n- Area: Subagents\n- Tier: Under the hood\n- Useful: 1\/5\n- Signal: 2\/5\n- Scope: individual\n- Heads-up: no"}