Copilot Chat Porter
A VS Code extension that enables GitHub Copilot Chat to export and import its own conversation history via Language Model Tools.
Features
Export Chat History
The AI agent can export the current chat session as a JSON file to the workspace — just ask in natural language:
- "把对话导出到 history/session1.json"
- "Export this conversation to exports/today.json"
- "保存当前对话" (auto-generates a timestamped filename)
Import Chat History
Restore a previously exported session and continue the conversation:
- "从 history/session1.json 恢复之前的对话"
- "Load the chat from exports/today.json"
The restored session opens in a new chat tab with full history.
What's in the exported JSON?
| Included |
Not Included |
| All user messages (raw text) |
Detailed tool call I/O |
| AI responses (including thinking) |
Per-turn token usage |
| File references and variables |
Raw LLM prompts |
| Agent and model info |
Debug-level logs |
| Timestamps |
|
| Edited file events |
|
Requirements
- VS Code
>= 1.99.0
- GitHub Copilot Chat extension
Usage
Via AI (recommended)
Simply tell Copilot in Agent Mode to export or import:
User: Save our conversation to history/chat.json
Agent: (calls exportChatHistory tool) → Done ✅
User: Restore the session from history/chat.json
Agent: (calls importChatHistory tool) → Opened in new tab ✅
You can also reference the tools explicitly with #exportChatHistory or #importChatHistory in the chat input.
Via Command Palette
The built-in VS Code commands also work independently:
Cmd+Shift+P → "Export Chat..." — save to any locationCmd+Shift+P → "Import Chat..." — restore from a file
Development
# Install dependencies
npm install
# Compile
npm run compile
# Watch mode
npm run watch
Press F5 in VS Code to launch the Extension Development Host for debugging.
How It Works
The extension registers two Language Model Tools that Copilot's agent can invoke automatically:
exportChatHistory — Calls VS Code's internal workbench.action.chat.export command with a workspace-relative path, bypassing the file dialog.importChatHistory — Calls workbench.action.chat.import to restore a session from a JSON file into a new chat tab.
The AI model reads each tool's inputSchema to understand the parameters, and extracts values from the user's natural language input (e.g., "save to exports/today.json" → { outputPath: "exports/today.json" }).
License
MIT
