{"version":"2.1.275","anchor":"forked-agent-runs-can-report-token-usage-per-request","canonical_anchor":"forked-agent-runs-can-report-token-usage-per-request","heading":"Forked agent runs can report token usage per-request","tier":"soon","area":"Subagents","scope":null,"heads_up":null,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.275\/e\/forked-agent-runs-can-report-token-usage-per-request","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.275","markdown":"### Forked agent runs can report token usage per-request\n\nForked agent runs can now report token usage as it streams in, not just at the end\n\n**What**\n\nWhen Claude Code runs a forked agent (a subagent branched off to do a separate task), it can now report token usage incrementally through a new `onRequestUsage` callback as responses stream in, in addition to the final total usage it already reported.\n\n**Why**\n\nThis lets code that manages forked agents track token consumption in near real time rather than waiting until the run finishes.\n\n- Area: Subagents\n- Names: `onRequestUsage`\n- Tier: Nothing to try yet\n- Useful: 3\/5\n- Signal: 2\/5"}