The artifacts/files tool's publish action now accepts a pin boolean parameter; when true, publishing also pins the artifact to the user's claude.ai sidebar, where it shows up the next time the sidebar loads. On success the tool reply includes a note that it was pinned to the sidebar; if pinning fails, the publish itself still succeeds and the reply appends a note that the pin didn't go through. This is wired into the publish tool's response builder at two call sites, gated by the tool's Zod schema still exposing a pin literal in its action shape.
Whether the field is even offered to the model is controlled by a check ($Ot()) that returns false outright for remote/cloud sessions (unless Jd()), otherwise reads the CLAUDE_CODE_ARTIFACT_PIN env var or falls back to a gate.
The entry above is what we published on the day. These lines were added later, as Anthropic's own pages caught up, and they sit beside the original rather than replacing it.
Without the `thinking-binding-controls-2026-08-01` beta header the message also names that header. Keep the conversation history append-only, or send the beta header with `prefix_mismatch_behavior: "drop_block"` to drop the block and conti…api/errors see the edit
| `claude-science uninstall` | Windows only. Remove the app, its shortcuts, and its PATH entry while keeping your data; `--purge` also deletes the data directory. Quit Claude Science first. |claude-science/command-line-settings see the edit
Anthropic's documentation has since written up pin, on Command line settings.