← 开源
carina-studio

ULogViewer

Cross-Platform, Agent-Native Universal Log Viewer.

ApplicationsSpecialistsC#
在 GitHub 打开
增长势头
+124 小时新增 Star+0.2%
597
Star
51
Fork
+0
本周
6
贡献者
创建于 2021-06-07 · 更新于 2026-10-16 · 今日第 6643 名
主要开发者
hamster620carina-studioylchu
README

ULogViewer

A cross-platform, agent-native log viewer for reading, parsing, and analyzing various types of logs. Please visit the Website for more details.

跨平台且支援 AI 代理的日誌檢視器,提供多種形式之日誌讀取、解析與分析。請參閱 網站 以取得更多資訊。

⚠️ NOTICE

If you encounter failure of upgrading from 2026.1.0 Preview, 2026.1.1 RC, or 2026.1.2 on Windows, please manually close all mcp.exe processes and try again.

📥 Download 2026.2

Windows

macOS

Linux

⭐ Supported Log Data Sources

A log data source reads raw log data into ULogViewer. It can be a file, a network stream, a database, or a script you write.

Category Log Data Sources
Local Standard Output (stdout)
Files
Windows Windows Event Log
Windows Event Log File
Network HTTP/HTTPS
TCP (without SSL)
UDP
Database SQLite Database
MySQL Database Pro
SQL Server Database Pro
Cloud Azure CLI Pro
Script Log Data Source Script Pro
Embedded Log Data Source Script Pro

Know more about

⭐ Log Profiles

Each log profile defines:

  • What log data source should be used.
  • How to parse log data into structured logs.
  • Which properties of the log should be displayed in the list.
  • Which properties of the log should be used to generate charts. Pro
  • How to output logs back to text (e.g., copying).

⭐ Log Filtering

Log filtering is one of the most important features in ULogViewer that helps you find and analyze problems from logs. You can filter logs by:

  • Text filter defined by a regular expression.
  • Level (Priority) of the log.
  • Process ID of the log, if available.
  • Thread ID of the log, if available.
  • Marked logs.

For text filter, you can also predefine some filters you may use frequently and filter logs by combination of these text filters.

Know more about

⭐ Log Marking

When viewing logs, you can mark some logs with different colors which are important for you. There is a separated side panel to list all marked logs to help you to jump to marked log quickly. Marked logs will be kept if you are viewing logs from files so that you don’t need to mark them again when you open log files next time.

⭐ Log Analysis

Except for log filtering, you can also define rule sets or write scripts to analyze logs. Log analysis runs in the background separately and generates results to a separate side panel. Currently there are 4 types of log analysis supported:

  • Key Log Analysis
    Find logs with a specific text pattern and level. You can extract information from the log and put it in the result message.
  • Operation Duration Analysis
    Find operations marked by specific starting and ending logs and calculate their duration. You can extract information from the log and put it in the result message.
  • Operation Counting Analysis
    Find operations marked with a specific text pattern and level, then count the number of operations that occurred for each given time frame. You can extract information from the log and put it in the result message.
  • Log Analysis Script
    Write script to analyze logs according to your requirement completely.

Know more about

⭐ AI Integration

Starting from 2026.1, ULogViewer Pro can expose your logs to AI clients through the Model Context Protocol (MCP). You can ask the AI assistant in your client to inspect logs, query tabs, set filters, mark logs and more. The following MCP clients have been verified to work with ULogViewer:

  • Claude Code
  • Claude Desktop
  • Codex
  • Cursor
  • Gemini CLI
  • LM Studio
  • Windsurf

MCP access is disabled by default. You can enable it in Options, where you can also control which actions (filtering, marking, selection, window activation) are allowed.

Logs are redacted by Sensitive Data Protection before they leave ULogViewer, masking content such as email addresses and IP addresses. It is applied to MCP responses by default, and its rules can be edited in Options.

🤝 Dependencies

排名相邻
同类项目