# Logfire — Python Observability on OpenTelemetry

> Logfire is Pydantic’s Python SDK for traces/metrics/logs on OpenTelemetry, helping teams add observability with minimal code and query data with SQL.

## Install

Save as a script file and run:

## Quick Use

1. Install:
   ```bash
   pip install logfire
   ```
2. Authenticate (creates local credentials):
   ```bash
   logfire auth
   ```
3. Verify:
   - Run the smallest example from the docs and confirm you can see at least 1 trace/span in the UI (or your chosen OTel backend).

## Intro

Logfire is Pydantic’s Python SDK for traces/metrics/logs on OpenTelemetry, helping teams add observability with minimal code and query data with SQL.

- **Best for:** Python teams who want quick, consistent observability for apps and agent backends
- **Works with:** Python apps (e.g., FastAPI) and OpenTelemetry-compatible backends
- **Setup time:** 10–30 minutes

## Practical Notes

- Quant: aim for 1 end-to-end request trace with spans for DB + HTTP within the first hour.
- Quant: track p95 latency and error rate before vs after adding instrumentation.

## Agent backend tip: make tool calls observable

If you run an agent service, the fastest win is instrumenting tool calls:
- external HTTP requests,
- database queries,
- vector search,
- and model API calls.

Name spans consistently (e.g., `tool.http`, `tool.sql`, `tool.vector`) so you can answer: *where did time and failures go?*

## Data hygiene

Avoid logging secrets. Configure scrubbing/redaction early and review what ends up in traces before enabling in production.

### FAQ

**Q: Is the whole platform open-source?**
A: No. This repo is the open-source SDK and docs; the hosted platform is separate.

**Q: Can I export to another backend?**
A: Yes—OpenTelemetry compatibility means you can target other OTel backends if needed.

**Q: What should I instrument first?**
A: Start with request traces + tool calls, then add DB/HTTP client instrumentation.

## Source & Thanks

> Source: https://github.com/pydantic/logfire
> License: MIT
> GitHub stars: 4,236 · forks: 234

---

<!-- ZH -->

## 快速使用

1. 安装：
   ```bash
   pip install logfire
   ```
2. 认证（生成本地凭据）：
   ```bash
   logfire auth
   ```
3. 验证：
   - 跑通文档的最小示例，确认至少 1 条 trace/span 出现在 UI（或你选用的 OTel 后端）。

## 简介

Logfire 是 Pydantic 团队提供的 Python SDK，用 OpenTelemetry 统一采集 traces/metrics/logs，并以更少的改动获得可观测性；适合把调试、性能与错误定位变成可查询的工程化数据。

- **适合谁：** 希望快速把可观测性落地到应用/agent 后端的 Python 团队
- **可搭配：** Python 应用（如 FastAPI）与任意兼容 OpenTelemetry 的后端
- **准备时间：** 10–30 分钟

## 实战建议

- 量化建议：1 小时内做到 1 条端到端请求 trace（至少含 DB + HTTP span）。
- 量化建议：对比接入前后 p95 延迟与错误率，观察收益与开销。

## Agent 后端提示：把工具调用纳入可观测性

跑 agent 服务时，最快的收益来自把“工具调用”打点：
- 外部 HTTP 请求、
- 数据库查询、
- 向量检索、
- 模型 API 调用。

给 span 统一命名（如 `tool.http`/`tool.sql`/`tool.vector`），才能快速回答“时间与失败都花在哪”。

## 数据卫生

避免把密钥写进日志/trace。尽早配置脱敏规则，并在生产启用前复核落盘内容。

### FAQ

**整个平台都开源吗？**
答：不是。这个仓库是开源 SDK 与文档；平台端是独立产品。

**能导出到别的后端吗？**
答：可以。兼容 OpenTelemetry，通常能接入其他 OTel 后端。

**优先打点哪些？**
答：先做请求链路 + 工具调用，再补齐 DB/HTTP 客户端等自动化集成。

## 来源与感谢

> Source: https://github.com/pydantic/logfire
> License: MIT
> GitHub stars: 4,236 · forks: 234


---
Source: https://tokrepo.com/en/workflows/logfire-python-observability-on-opentelemetry
Author: Script Depot