MCP.so
ログイン
C

Chronicle MCP Server

@Leviathan0x0

基本情報

カテゴリ

AI とエージェント

トランスポート

stdio

作者

Leviathan0x0

投稿者

Leviathan

Chronicle MCP Server について

Chronicle MCP is a high-performance, universal local chat history archive connector designed specifically for Model Context Protocol (MCP) clients like Cursor, Claude Code, and daily development environments. Built natively in Python, it serves as an intelligent local storage and

Chronicle MCP Server の設定

以下の設定を使って、このサーバーを MCP 対応クライアントに追加してください。

{
  "mcpServers": {
    "chronicle-mcp": {
      "command": "uvx",
      "args": [
        "chronicle-mcp-server"
      ]
    }
  }
}

Chronicle MCP Server のツール

ツール一覧はまだありません

この掲載には表示できるツール一覧がまだありません。サーバーにツールがないという意味ではありません。対応機能はプロジェクトのドキュメントをご確認ください。

上のボタンで一時的なサンドボックスでこのサーバーを実行し、ツール一覧を取得できます。

概要

What is Chronicle MCP?

Chronicle MCP is a production-grade Model Context Protocol server that syncs, cleans, formats, and indexes local artificial intelligence chat transcripts. It allows agents to search, compare, retrieve, and reference past conversation logs while using optimized token-saving heuristics that can reduce context window utilization by up to 40%.

How to use Chronicle MCP?

Install the server and run chronicle to launch the stdio transport for MCP clients. Use chronicle add to configure IDE integrations (Cursor, Claude Code, VS Code, Trae) with automatic path resolution, and chronicle split to break monolithic chat exports into individual files. Customize the storage folder with --chats-folder.

Key features of Chronicle MCP

  • Format normalization engine for ChatGPT, Claude, generic JSON, and Markdown.
  • Context token optimization with code block summarization and length limiting.
  • Six versatile tools: search, read, sync, compile insights, maintain storage.
  • Cross-platform path resolution for macOS, Windows, and Linux.
  • Native IDE integration with fallback engine for emerging platforms.
  • Structural split engine for large conversation JSON exports.

Use cases of Chronicle MCP

  • Search and filter local chat transcripts by keyword, semantic similarity, or date range.
  • Import and sync conversation exports from ChatGPT or Claude into a local archive.
  • Compare two chat transcripts to identify shared and unique terms.
  • Extract action items, task lists, and checkboxes from chat logs.
  • Compile a project brief by synthesizing multiple chat summaries into one document.

よくある質問

1

Which chat export formats are supported?

Chronicle normalizes exports from ChatGPT (nested mapping nodes), Claude (flat chat_messages lists), and generic JSON or Markdown logs from tools like Cline or Continue.

2

How does token optimization work?

It automatically replaces verbose code blocks with metadata summaries (language and line count) and truncates extremely long messages at a configurable character threshold.

3

How can I split a large conversation file?

Use the `chronicle split` subcommand followed by the path to the JSON file and an output directory, e.g., `chronicle split /path/to/conversations.json --out /path/to/output_directory`.

4

How do I configure a custom storage folder?

Run `chronicle --chats-folder /path/to/custom/chats` to save a custom global folder path in `~/.chronicle_settings.json`.

5

Which IDEs are supported out of the box?

Chronicle directly integrates with Cursor, Claude Code, VS Code (Cline/RooCode/Continue), and Trae, with a dynamic fallback engine for other platforms.

コメント

「AI とエージェント」の他のコンテンツ