What this video shows
The presenter pins one chat as a coordinator and asks it to inspect a set of 25 existing chats. In the recorded test, it found the two chats waiting for an answer, opened four new chats and checked the resulting work. The video also shows sections, scheduled morning checks, projects, side chats, split view, phone access, document and slide work, dashboards and undo.
Treat the demonstration as one user's product test rather than a reliability study. Start with read-only status checks, inspect which chats and files Claude can access, and keep human review before messages, documents or external actions leave the workspace.
The video's YouTube page and chapters provide the title, creator, October 28 upload date, duration and feature sequence. Use Learnetto's agent test-set guide to turn a convenience workflow into repeatable checks.
What you will learn
- Use a coordinator chat for status and routing before allowing it to create or change work.
- Separate projects by data and purpose so cross-chat access stays understandable.
- Review scheduled outputs and access scope after every product update.
- Keep a person responsible for messages, shared documents and consequential actions.
How to apply this safely
- Create five disposable chats with two clearly unanswered questions.
- Ask one pinned chat to report only what is waiting, then compare its list with the known state.
- Test one scheduled read-only check and inspect which chats, projects and files it can access.
- Add document or dashboard work only after the status test passes and a human review step is explicit.
Important limitations
- The creator conducted one 25-chat test, so the result does not establish reliability across accounts, plans or future app versions.
- The video is an independent tutorial and does not replace Anthropic's current product documentation.
- Several features can expose information across chats or projects if access boundaries are wider than expected.
- An undo control may reverse some edits, but it cannot recall every message or external action.
- The creator promotes free and paid workflow guides in the description.
Sources to check
- Claude Can Now Run All Your Other Chats For You Hyperautomation Labs video published October 28, 2026.
- Build an AI agent test set Learnetto's workflow for known cases and release thresholds.
Continue learning on Learnetto
Build an AI agent test set
Create stable cases for cross-chat status and routing.
Sandbox AI coding agents
Limit permissions and external actions.
Review AI-generated code
Apply evidence-led human review to generated work.