{"version":"2.1.246","anchor":"cloud-agent-output-is-sanitized-before-use","canonical_anchor":"cloud-agent-output-is-sanitized-before-use","heading":"Cloud agent output is sanitized before use","tier":"internal","area":"Cloud Sessions","scope":null,"heads_up":null,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.246\/e\/cloud-agent-output-is-sanitized-before-use","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.246","markdown":"### Cloud agent output is sanitized before use\n\nText coming back from a remote-isolation agent is cleaned before you see it or it's logged.\n\n**What**\n\nText returned by a cloud session in the workflow `agent({isolation:'remote'})` path is now sanitized, and the sanitized form is what gets returned to you, logged and passed to the handback classifier, rather than the raw text.\n\n**Evidence**\n\n`agent({isolation:'remote', schema}) completed without structured output: `\n\n- Area: Cloud Sessions\n- Tier: Under the hood\n- Useful: 2\/5\n- Signal: 3\/5"}