{"version":"2.1.317","anchor":"artifact-tool-marks-read-results-with-a-prefix-on-the-resul","canonical_anchor":"artifact-tool-marks-read-results-with-a-prefix-on-the-resul","heading":"Artifact and docs reads can now be marked with markReads","tier":"internal","area":"Artifacts","scope":"individual","heads_up":false,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.317\/e\/artifact-tool-marks-read-results-with-a-prefix-on-the-resul","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.317","markdown":"### Artifact and docs reads can now be marked with markReads\n\nWhen an artifact or doc has `markReads` set, Claude Code marks what Claude reads from it and prefixes those tool results with a marker\n\n**Unclear.** What the marker text says and what turns marking on for an artifact are not shown.\n\n**What**\n\nArtifacts are documents, apps and data Claude can create and read through the artifact tool. A new `markReads` setting on a stored artifact or doc changes how its reads are reported:\n\n- The artifact tool actions `read`, `list_files`, `read_file`, `list_assets`, `read_asset`, `read_db`, `comments`, `verify`, `get_endpoints`, `call_endpoint` and `run_script` now record their output and the id of the call as marked reads when the artifact's stored record has `markReads` set.\n\n- When an output is marked, the result Claude sees starts with a marker text. What the text says, and what sets `markReads`, were not traced.\n\n- A docs read result now returns `markReads: true` when the flag is set, instead of `authoredByOthers`.\n\n- An artifact database read does the same: `markReads` when the flag is set, otherwise `authoredByOthers` as before.\n\n**Why**\n\nContent read from some artifacts and docs is now labelled in what Claude sees, in place of the earlier `authoredByOthers` marking for those cases. What the label is used for is not shown.\n\n- Area: Artifacts\n- Tier: Under the hood\n- Useful: 1\/5\n- Signal: 3\/5\n- Scope: individual\n- Heads-up: no"}