Jtree
Display Jaeger traces as a hierarchical tree in your terminal
Install / Use
/learn @tomarrell/JtreeREADME
🌲 jtree
Display Jaeger traces as a hierarchical tree in your terminal. The compact output is ideal for piping to LLM CLI agents, enabling AI-assisted trace analysis without overwhelming context.
Usage
# Basic usage
jtree <trace-id>
# Pass a full Jaeger URL directly
jtree http://localhost:16686/trace/abc123def456
# Filter to slow error spans
jtree -min-duration 100ms -error <trace-id>
# Verbose JSON output with all tags
jtree -json <trace-id>
LLM Integration
Pipe trace data directly to LLM CLI agents for AI-assisted debugging:
jtree abc123 | claude "What's causing the latency in this trace?"
jtree -error abc123 | claude "Explain these errors and suggest fixes"
Output
Default human-readable format:
call-abc123 [orchestrator] 16:43:33.529 55.47s
stt.websocket.connect [orchestrator] 16:43:33.539 810.76ms
conversation.turn.bot [orchestrator] 16:43:41.178 3.11s
tts.turn [orchestrator] 16:43:41.179 3.11s
tts.reader [orchestrator] 16:43:41.179 3.11s
With -relative:
call-abc123 [orchestrator] +0us 55.47s
stt.websocket.connect [orchestrator] +9.94ms 810.76ms
conversation.turn.bot [orchestrator] +7.65s 3.11s
JSON format (-json):
call-abc123 {"duration":"55.47s","service":"orchestrator","span_id":"f1f173a9f8639951","tags":{...}}
Installation
Homebrew (macOS/Linux)
brew install tomarrell/tap/jtree
Go
go install github.com/tomarrell/jtree@latest
Binary
Download the latest binary from the releases page.
Flags
| Flag | Default | Description |
|------|---------|-------------|
| -url | http://localhost:16686 | Jaeger URL |
| -json | false | Output verbose JSON with all tags |
| -min-duration | 0 | Only show spans with duration >= value (e.g. 100ms, 1s) |
| -error | false | Only show error spans and their ancestors |
| -service | | Only show spans from this service |
| -depth | 0 | Limit tree depth (0 = unlimited) |
| -relative | false | Show timestamps relative to trace start |
| -version | false | Print version and exit |
License
MIT
Related Skills
node-connect
344.4kDiagnose OpenClaw node connection and pairing failures for Android, iOS, and macOS companion apps
frontend-design
99.2kCreate distinctive, production-grade frontend interfaces with high design quality. Use this skill when the user asks to build web components, pages, or applications. Generates creative, polished code that avoids generic AI aesthetics.
openai-whisper-api
344.4kTranscribe audio via OpenAI Audio Transcriptions API (Whisper).
qqbot-media
344.4kQQBot 富媒体收发能力。使用 <qqmedia> 标签,系统根据文件扩展名自动识别类型(图片/语音/视频/文件)。
