# Monoscope — LLM Query for Logs/Traces/Metrics

> Monoscope stores logs/traces/metrics in S3-compatible buckets and lets you explore them with natural-language queries plus a CLI and self-hosted UI.

## Install

Save as a script file and run:

## Quick Use

```bash
git clone https://github.com/monoscope-tech/monoscope.git
cd monoscope
docker-compose up
```

Visit `http://localhost:8080` (default: admin/changeme)

```bash
monoscope auth login
monoscope send-event -m "Hello from Monoscope"
```

## Intro

Monoscope stores logs/traces/metrics in S3-compatible buckets and lets you explore them with natural-language queries plus a CLI and self-hosted UI.

- **Best for:** Log/trace triage and NLQ for observability data
- **Works with:** Docker Compose; S3-compatible storage; OpenTelemetry ecosystem
- **Setup time:** 10–30 minutes

## Practical Notes

- GitHub: 692 stars · 23 forks; pushed 2026-05-11 (verified via GitHub API).
- README quick start uses `docker-compose up` and the UI defaults to `http://localhost:8080`.
- Per README: CLI emits a stable JSON envelope, making it easy to chain search → triage in agent workflows.

## Main

A “minimum useful loop” for incidents:

1. Ingest a small amount of real telemetry (or use the test send/telemetrygen commands).
2. Ask a narrow question first (“show errors for service X”), then progressively widen the window.
3. Once you find a suspicious trace/log event, keep the **query trail** (filters + IDs) as a repeatable runbook.

When you later connect an agent, the goal is not “one-shot answers”—it’s to have the agent produce a deterministic chain of CLI commands you can re-run and audit.

### FAQ

**Q: Is cloud required?**
A: No—README includes a self-host quick start with Docker Compose.

**Q: Where is data stored?**
A: README states it stores telemetry in S3-compatible buckets.

**Q: How do I make it agent-friendly?**
A: Prefer CLI-driven workflows and keep the command chain as your audit trail.

## Source & Thanks

> Source: https://github.com/monoscope-tech/monoscope
> License: AGPL-3.0
> GitHub stars: 692 · forks: 23

---

<!-- ZH -->

## 快速使用

```bash
git clone https://github.com/monoscope-tech/monoscope.git
cd monoscope
docker-compose up
```

访问 `http://localhost:8080`（默认：admin/changeme）

```bash
monoscope auth login
monoscope send-event -m "Hello from Monoscope"
```

## 简介

Monoscope 将日志/追踪/指标存入 S3 兼容对象存储，并提供 Web UI + CLI：你可以像查日志一样用自然语言与结构化过滤定位异常，还能把查询链路交给 agent 自动化（自托管或云端）。

- **适合谁：** 日志/追踪排障 + 用自然语言查询观测数据
- **可搭配：** Docker Compose；S3 兼容对象存储；OpenTelemetry 生态
- **准备时间：** 10–30 分钟

## 实战建议

- GitHub：692 stars · 23 forks；最近更新 2026-05-11（GitHub API 验证）。
- README quick start 直接用 `docker-compose up`，并给出 UI 默认地址 `http://localhost:8080`。
- README：CLI 输出稳定的 JSON envelope，适合让 agent 串起“发现 → 搜索 → 定位”的链路。

## 主要内容

一个“最小可用”的排障闭环：

1. 先接入少量真实观测数据（或用 README 的测试命令造数据）。
2. 先问窄问题（例如“服务 X 的 error”），再逐步扩大时间窗口与相关维度。
3. 找到可疑 trace/log 后，把 **查询轨迹**（过滤条件 + 关键 ID）沉淀成可复跑的 runbook。

后续接入 agent 时，目标不是“一次性回答”，而是让 agent 输出一串确定性的 CLI 命令，便于复现与审计。

### FAQ

**必须用云端吗？**
答：不必须。README 提供 Docker Compose 自托管 quick start。

**数据存哪？**
答：README 表示将观测数据存入 S3 兼容对象存储。

**怎么做成 agent 友好？**
答：尽量用 CLI 驱动，把命令链当作审计与复现依据。

## 来源与感谢

> Source: https://github.com/monoscope-tech/monoscope
> License: AGPL-3.0
> GitHub stars: 692 · forks: 23


---
Source: https://tokrepo.com/en/workflows/monoscope-llm-query-for-logs-traces-metrics
Author: Script Depot