Copilot Export: Save Useful Work for Your Next Assistant

A Copilot export follows the app and account: personal Microsoft accounts have an activity-history download, while GitHub Copilot chat in VS Code has its own export command. Work or school accounts use a separate organizational process, so the personal-account instructions do not apply everywhere. Save the record you need, then check it before handing the work to another assistant.

fenbs can hold the decisions and progress you manually select from that record, so the next assistant has a useful starting point. It does not import Copilot chat history automatically. The goal is a checked account of what matters next, with the original material kept separately when you need it.

Which Copilot export route should you use?

1. Identify the app and signed-in account where the conversation happened. Write down whether it was personal Microsoft Copilot, a work or school Microsoft 365 experience, or GitHub Copilot in an editor.
2. Choose the matching route below. An export option in one product is not evidence that another Copilot product has the same menu.
3. Open the saved material and check the conversation you actually need. Keep the approved output, supporting sources, and remaining work together in a short handoff.

If you are also choosing where to continue the work, Copilot vs ChatGPT covers the broader product comparison. This guide focuses on saving and checking the work you already have.

How do you export personal Microsoft Copilot history?

For a personal Microsoft account, follow Microsoft’s activity-history export instructions↗ (opens in a new tab). Sign in to the Microsoft privacy dashboard with the account you used for the conversation.

1. Open Privacy, then Empower your productivity, then Copilot and Your Copilot app activity history.
2. Choose Copilot apps for that activity, then Export all activity history. The download is a CSV, a table stored as a text file that you can open in a spreadsheet app.
3. For activity in Microsoft 365 apps under the same personal account, choose Copilot in Microsoft 365 apps instead. Select the export for the app activity or Copilot Chat activity you need.

Keep the original download unchanged and inspect a working copy. Look for a conversation you recognize, then compare a useful answer with the chat still visible in the app. Check any source documents separately; a record of a discussion is not a substitute for the document you intend to deliver.

What changes with a work or school account?

The personal privacy-dashboard export instructions explicitly exclude work and school accounts. For organizational Microsoft Copilot, Microsoft documents administrative access through Content search or Microsoft Purview in its data and privacy guidance↗ (opens in a new tab). That is a separate process from downloading personal activity history.

Ask your organization’s administrator which approved route applies when you need a formal record. Explain the app, account, conversation, and purpose. If you only need to continue a task, check whether copying the relevant output into an approved work document meets that need before requesting a broader history export.

Keep the handoff within the places your organization permits. For example, an internal proposal may need a summary of the approved outline and a link to the controlled document, without copying client details into another service. Exporting something does not decide who should receive it.

How do you export GitHub Copilot chat in VS Code?

In VS Code, open the chat session, open the Command Palette, and run Chat: Export Chat.... Choose where to save the JSON file, a structured text record containing the prompts and responses. VS Code’s session documentation↗ (opens in a new tab) also describes Copy and Copy All for copying messages as Markdown, plain text with formatting marks.

Those are VS Code instructions. Check the relevant documentation if you used GitHub Copilot in another editor or interface. Do not assume the same command exists in the desktop app or on GitHub’s website.

Preserve the actual code changes separately. A transcript that says a file was fixed is not evidence that the saved file contains the fix or that a test passed. Include the relevant file or commit reference and the checks you ran. For the wider question of what an editor session should receive, see context engineering with GitHub Copilot.

What should a checked handoff include?

Read the selected conversation from beginning to end before summarizing it. Early suggestions can become obsolete later in a chat. Separate the assistant’s proposals from choices a person approved, and distinguish an intended action from something actually completed.

• Goal: what the work is trying to produce and where the current output lives.
• Decisions: what a person chose, who chose it, and which constraints still apply.
• Progress: what changed, what was checked, and the result of those checks.
• Uncertainty: missing sources, unresolved questions, or work waiting on a person.
• Next step: one concrete action the next person or assistant can start.

Say a small design studio used Copilot to draft an installation guide. A useful handoff says the owner approved the short outline, the draft is saved in the project folder, and the safety section still needs review. “Copilot finished the guide” would hide the most important remaining task.

Keep confirmed choices in fenbs Decisions and rules, with open questions left unresolved until a person decides them.

How to do this with fenbs?

1. Create a free account, then add a project for the work you are handing over.
2. Add the remaining tasks and their plans. Manually record the decisions a person approved, including the reason and any continuing rules.
3. Save project progress in Where we left off. State the current output, completed checks, unresolved questions, and next action.
4. Use the integration guides to connect a supported assistant. GitHub Copilot is supported; do not assume a Microsoft Copilot conversation has the same connection. Ask the connected assistant to read the project context and explain its starting point.

For a fuller transition checklist, use the AI handoff guide. fenbs is not the right fit if you only need an untouched archive of every conversation; keep that export in suitable file storage. Its job here is the active work you deliberately carried forward.

What if the export is missing something you need?

First check the account, app, and chosen export category against the conversation you expected. Then inspect the original conversation and source files directly. Write down any missing material instead of asking another assistant to invent a reconstruction.

A shared link serves a different purpose from a downloaded record. Microsoft’s conversation-history guidance↗ (opens in a new tab) says personal Copilot conversation links show a snapshot and can be opened by anyone who has the link. Review the preview before using one as a reference.

For the neighboring workflows, see Copilot for writing, Copilot documentation work, and exporting notebook research. The AI handoffs hub collects ways to preserve useful work when the next session starts elsewhere.