Copilot Prompt Logger

A VS Code extension that provides a @logger chat participant for logging GitHub Copilot conversations. Capture and review your AI interactions with timestamps, model info, and user details in persistent JSON files.

🚀 Features

Chat Participant (@logger)

• Custom chat participant that integrates directly with GitHub Copilot Chat
• Persistent logging - all logs are saved to file and survive VS Code restarts
• Append-only storage - never overwrites previous logs
• Chat history capture - retroactively log previous conversations
• Circular buffer - configurable memory buffer for recent entries
• JSON export - structured data for analysis and backup

Commands Available

• /view - Display recent log entries
• /export - Export logs as formatted JSON
• /clear - Clear all logs (memory and file)
• /history - Log previous chat conversations from current session
• /help - Show all available commands

Manual Logging

• Command palette integration for manual prompt logging
• Configurable log file name and buffer size
• Workspace-specific storage in .vscode folder

📖 How to Use

Using @logger in Copilot Chat

1. Basic Logging: Mention @logger with any message to log it

   @logger How do I create a React component?
   

2. View Recent Logs:

   @logger /view
   

3. Export Logs as JSON:

   @logger /export
   

4. Log Previous Conversation History:

   @logger /history
   

   This captures all previous messages from your current chat session.

5. Clear All Logs:

   @logger /clear
   

6. Get Help:

   @logger /help
   

Manual Logging via Command Palette

1. Open Command Palette (Ctrl+Shift+P / Cmd+Shift+P)
2. Type Capture Prompt Log and press Enter
3. Enter your prompt and model name when prompted
4. Use Show Prompt Log to view entries

📁 Log File Location

Logs are stored in your workspace's .vscode folder as prompt-log.jsonl by default.

Example log file content:

{"prompt":"How do I create a React component?","model":"user-request","timestamp":"2025-09-07T12:34:56.789Z","user":"john"}
{"prompt":"To create a React component, you can use...","model":"copilot-response","timestamp":"2025-09-07T12:35:02.123Z","user":"john"}
{"prompt":"Can you show me hooks?","model":"user-request","timestamp":"2025-09-07T12:36:15.456Z","user":"john"}

⚙️ Configuration

Customize the extension via VS Code settings:

{
  "copilotPromptLogger.logFileName": "my-prompt-log.jsonl",
  "copilotPromptLogger.maxLogEntries": 2000
}

Available Settings:

• Log File Name: Change the name of the log file (default: prompt-log.jsonl)
• Max Log Entries: Set circular buffer size for memory (default: 1000)

🔧 Technical Details

What Gets Logged:

• User prompts sent to GitHub Copilot
• Copilot responses (markdown, file trees, anchors, commands)
• Timestamp (ISO 8601 format)
• Model information (when available)
• User details (system username)

Storage:

• Persistent file storage in workspace .vscode folder
• Circular buffer in memory for performance
• Append-only logging - never loses data
• JSONL format - one JSON object per line for easy parsing

Limitations:

• Cannot passively monitor native Copilot conversations
• Requires explicit @logger mention to capture chat
• Only logs when users interact with @logger participant
• Workspace-specific logging (not global)

🚦 Getting Started

1. Install the extension from VS Code Marketplace
2. Open a workspace in VS Code
3. Start Copilot Chat and type @logger /help to see commands
4. Begin logging by mentioning @logger with your prompts

💡 Example Workflow

# Start a conversation with Copilot
You: How do I handle errors in React?
Copilot: You can handle errors in React using...

# Later, log the conversation
You: @logger /history
Extension: Previous chat history logged successfully.

# View what was logged
You: @logger /view
Extension: [Shows formatted log entries]

# Export for external analysis
You: @logger /export
Extension: [Shows JSON formatted logs]

🔒 Privacy & Security

• Local storage only - logs stay in your workspace
• No external transmission - data never leaves your machine
• User-controlled - you decide what to log and when
• Transparent - all logged data is viewable and exportable

🛠️ Development

Built with:

• TypeScript for type safety
• VS Code Extension API for integration
• Circular buffer for efficient memory usage
• File system APIs for persistent storage

Enjoy logging your AI conversations! 🎉t Logger

Easily capture and review prompts, model names, timestamps, and user info directly in Visual Studio Code.

Features

• Log prompts, model, timestamp, and user with a simple command
• Circular log file (oldest entries are removed automatically)
• View the last 10 log entries at any time

How to Use

1. Open the Command Palette (Ctrl+Shift+P or Cmd+Shift+P)
2. Type and select Capture Prompt Log to add a new entry
   • Enter your prompt and model name when prompted
3. To review logs, run Show Prompt Log from the Command Palette
4. The log file (prompt-log.jsonl by default) is saved in your workspace’s .vscode folder

Log File Format

Each log entry is a single line in JSON format, for example:

{"prompt":"example prompt","model":"gpt-4","timestamp":"2025-09-06T12:34:56.789Z","user":"yourname"}

Configuration

You can change the log file name and the maximum number of log entries in your VS Code settings:

• Copilot Prompt Logger: Log File Name (default: prompt-log.jsonl)
• Copilot Prompt Logger: Max Log Entries (default: 1000)

Requirements

No special requirements. Just install and use!

Support

If you have questions or feedback, please use the Q&A or Issues section in the Marketplace.

Thank you for using Copilot Prompt Logger!