AgentSkillsCN

laminar-instrument-codebase

为现有代码库接入Laminar追踪功能：选择需要观测的函数，正确初始化Laminar，添加标签/元数据/会话ID，并验证追踪数据。当用户提出希望添加Laminar追踪、为函数注入追踪功能，或希望将Laminar集成到TS/JS或Python代码库中时，可选用此方法。

SKILL.md
--- frontmatter
name: laminar-instrument-codebase
description: "Instrument an existing codebase with Laminar tracing: choose which functions to observe, initialize Laminar correctly, add tags/metadata/session IDs, and verify traces. Use when a user asks to add Laminar tracing, instrument functions, or integrate Laminar with a TS/JS or Python codebase."

Laminar Instrument Codebase

Workflow

  1. •Identify runtime and entrypoints; map critical flows and select functions to instrument.
  2. •Initialize Laminar once, early in the app, and configure auto-instrumentation for the libraries in use.
  3. •Add manual spans with observe around selected functions; attach tags, metadata, session/user IDs; suppress sensitive inputs/outputs.
  4. •Run a representative flow, validate traces in the UI, and tune span granularity and tags.

References

  • •references/function-selection.md for heuristics on what to instrument.
  • •references/ts-instrumentation.md for TypeScript/JavaScript patterns and snippets.
  • •references/python-instrumentation.md for Python patterns and snippets.