Parseur

io.github.parseur由 GitHub 账号 parseur 发布v1.6.0更新于 Nov 3, 2026

Manage Parseur.com mailboxes, parser fields, documents, exports and webhooks.

概览

AI 生成的概览

本地 MCP 服务器,让助手通过 Parseur API 管理 Parseur 邮箱、解析字段、文档、导出和 Webhook。

可以做什么
  • 列出、搜索、排序并查看邮箱,包括邮箱详情和解析结构
  • 列出、筛选并查看文档,含详情、处理日志和已提取的解析字段
  • 按文件、文件夹或文本上传文档,并重新处理、跳过、复制、拆分或删除文档
  • 将邮箱、字段或导出结果下载为 CSV、JSON 或 XLSX
  • 创建、列出、启用、暂停和删除 Webhook,并监听文档与表格事件
适用场景
适合已使用 Parseur 做邮件或文档数据提取的用户,希望助手在对话中查看邮箱、核对解析结果、上传新文档或配置 Webhook 集成时使用。
运行要求
以 Python 包 parseur-py 通过 uvx 在本地运行。需要 Parseur 账号和 API 密钥,通过环境变量 PARSEUR_API_KEY 提供;可用 PARSEUR_API_BASE 覆盖 API 基础地址。需要能访问 Parseur API 的网络。
安装前请注意
API 密钥可访问你的 Parseur 账号数据,请勿写入共享配置或日志。上传、重新处理、跳过、复制、拆分和删除等文档操作会改动真实数据,Webhook 的创建与删除会影响线上集成。可选的实时监听功能会经第三方 localtunnel 服务器转发数据,不受 Parseur 隐私条款约束,且没有端到端加密。

安装

在 SourceWeft 中

  1. 打开 控制台中的 Parseur,将其添加到工作区。
  2. 为需要使用其工具的对话启用该服务。

Desktop only,通过 STDIO。 STDIO 服务会启动本地进程,因此需要 SourceWeft 桌面宿主。

其他 MCP 客户端

参照 仓库 中的启动说明。

README

🤖🧙parseur-py

parseur-py is a modern Python client for the Parseur API.

It lets you manage mailboxes, documents, uploads, and webhooks programmatically or from the command line.

Built to help you automate document parsing at scale, parseur-py makes integrating with Parseur fast, easy, and Pythonic.

[GitHub Repo] [PyPI version] [License: MIT] [Read the Docs] [PyPI Downloads]


✨ Features

✅ List, search, and sort mailboxes
✅ Get mailbox details and schema
✅ List, search, filter, and sort documents
✅ Upload documents by file or email content
✅ Reprocess, skip, copy, or delete documents
✅ Manage custom webhooks for real-time events
✅ Listen to events in real time with a temporary webhook & tunnel
✅ Fully-featured Command Line Interface (CLI)


⚠️ Disclaimer about Localtunnel

When using the parseur listen command (with event listener support), your data is forwarded through localtunnel servers.

These servers are not affiliated with Parseur and are not covered by Parseur’s Privacy Policy or Data Processing Agreement.

Data transmitted through localtunnel is not encrypted end-to-end.

➡️ Use this feature at your own risk.

For production-grade setups, we strongly recommend configuring your own secure webhook endpoint instead of relying on localtunnel.


🚀 Quick Start

Install the package

bash
pip install parseur-py

With event listener support (Flask + localtunnel)

bash
pip install "parseur-py[listener]"

Install the package from source

bash
pip install -e .

Build documentation

bash
pip install -r requirements-doc.txtcd docsmake html

Run the tests

Unit tests run fully offline:

bash
pytest

Integration tests hit the real Parseur API. They create and delete real resources, so they're skipped unless credentials are provided via the environment (never committed or stored in ~/.parseur.conf):

bash
PARSEUR_API_BASE=https://api.parseur.com \PARSEUR_API_KEY=sk_your_key \pytest tests/integration -v

Initialize your configuration

Store your Parseur API credentials securely:

bash
parseur init --api-key YOUR_PARSEUR_API_KEY

Your config is saved (by default) in:

~/.parseur.conf

Example usage

List all your mailboxes:

bash
parseur list-mailboxes

List documents in a mailbox:

bash
parseur list-documents 12345

Upload a file to a mailbox (add --wait to block until it is parsed, with a live progress bar):

bash
parseur upload-file 12345 ./path/to/document.pdfparseur upload-file 12345 ./path/to/document.pdf --wait

Download a mailbox's results as a file (stdout by default, or --output):

bash
parseur download-mailbox 12345 --format csv -o results.csv     # whole mailboxparseur list-parser-fields 12345                               # find a table field idparseur download-field 12345 PF951 --format xlsx -o lines.xlsx  # a table field

Register a custom webhook:

bash
parseur create-webhook --event document.processed --target-url https://yourserver.com/webhook --mailbox-id 12345

Listen to events in real time (requires [listener]):

bash
parseur listen --event document.processed --mailbox-id 12345

With forwarding:

bash
parseur listen --event document.processed --mailbox-id 12345 --redirect-url http://localhost --redirect-port 8000

📜 CLI Commands

Run:

bash
parseur --help

for a full list of available commands.

Highlights

  • init: Set your API token and (optional) base URL
  • list-mailboxes: Search and sort mailboxes
  • get-mailbox: Fetch a mailbox by ID
  • get-mailbox-schema: Get the mailbox parsing schema
  • list-parser-fields: List the fields extracted by a mailbox
  • download-mailbox / download-field / download-export: Download results as a file (csv/json/xlsx)
  • list-export-configs: List a mailbox's custom export configurations
  • list-documents: Advanced document search, filtering, sorting
  • get-document / get-document-logs: Fetch document details and processing logs
  • reprocess-document / skip-document / copy-document / split-document / reverse-split-document / delete-document: Document lifecycle operations
  • upload-file / upload-text: Upload new documents (add --wait for synchronous parsing)
  • upload-folder: Upload every file matching a glob path
  • create-webhook / get-webhook / list-webhooks / delete-webhook: Create, get, list, and delete custom webhook integrations.
  • enable-webhook / pause-webhook: Activate or pause a webhook for a specific mailbox.
  • listen: Create a temporary webhook and listen to events in real time (with optional redirect & silent mode)

🔎 Advanced Search & Filtering

Mailbox listing supports:

  • Search by name or email prefix
  • Sort by:
    • name
    • document_count
    • template_count
    • PARSEDOK_count (processed)
    • PARSEDKO_count (failed)
    • QUOTAEXC_count (quota exceeded)
    • EXPORTKO_count (export failed)

Document listing supports:

  • Search in:
    • document ID
    • document name
    • template name
    • email addresses (from, to, cc, bcc)
    • document metadata header
  • Sort by:
    • name
    • created (received date)
    • processed date
    • status
  • Filter by:
    • received_after / received_before dates
  • Include parsed result in response

⚡ Webhooks Support

Easily register custom webhooks for events like:

  • document.processed
  • document.processed.flattened
  • document.template_needed
  • document.export_failed
  • table.processed
  • table.processed.flattened

Your webhook endpoint will receive POST notifications with Parseur payloads, enabling real-time integrations with your systems.


🤖 MCP Server (AI assistants)

Parseur hosts an MCP server at https://api.parseur.com/mcp. Add this URL to any MCP-compatible AI assistant (Claude, Codex, Cursor, …) to drive Parseur from it. No local install is needed.


🛠️ Configuration

Your API token and settings are stored in a simple INI file:

[parseur]api_token = YOUR_API_KEYbase_url = https://api.parseur.com

You can customize the path by setting `--config-path` in your calls if needed.


🐍 Python Client Usage

Beyond the CLI, parseur-py is a standard Python library. Example:

python
import parseur
parseur.api_key = "YOUR_API_KEY"
for mailbox in parseur.Mailbox.list():    print(mailbox.name)

Per-call API key override

Every method accepts an optional api_key argument that takes priority over the global parseur.api_key for that single call — useful for multi-account or multi-tenant code:

python
parseur.Mailbox.list(api_key="sk_account_a")parseur.Document.upload_file(123, "invoice.pdf", api_key="sk_account_b")

📖 Documentation

  • Parseur Official API Docs
  • This package mirrors Parseur’s REST API, adding pagination handling, schema support, and convenient CLI commands.

💼 License

MIT License


🤝 Contributing

We welcome contributions! Please:

  1. Fork the repo
  2. Create your feature branch (git checkout -b feature/foo)
  3. Commit your changes (git commit -am 'Add foo')
  4. Push to the branch (git push origin feature/foo)
  5. Open a pull request

✨ Credits

Developed with ❤️ by the Parseur team.


Parseur is the easiest way to automatically extract data from emails and documents. Stop copy-pasting data and automate your workflows!

来源:README.md,提交 e6c4aab

工具

0
SourceWeft 尚未在沙箱中启动过该软件包,因此其工具未知。

版本历史

1
  1. v1.6.0最新Sep 16, 2026

评分

已满足适用检查项的 50%

该评分衡量此收录条目的完整程度,依据注册中心条目、端点与公开来源计算得出。它不是对服务器本身的评价,也不表示服务器安全或当前可用:可连通性是带日期的检查结果,评分也不包含安装量或用户评价。

  • 近 30 天内可连通25 分

    请保持端点能够响应 MCP 握手;SourceWeft 每 30 天探测一次。

  • 已声明认证方式10 分

    请在注册中心条目中声明凭据请求头,或保持端点能够响应探测,以便判定其认证方式。

  • 已列出工具15 分

    请让服务器在无凭据时响应 tools/list,或发布可在沙箱中无凭据启动的软件包。

  • 描述必需15 分
  • 图标10 分
  • 文档15 分
  • 许可证10 分