tracekit-apm-setup

Installation
SKILL.md

TraceKit APM Setup

Auth Bootstrap

Do not tell the user to go sign up, log in separately, or manually create an API key before setup. First check for existing TraceKit auth. If ~/.tracekitconfig does not contain the production profile or TRACEKIT_API_KEY is missing, apply the tracekit-auth skill first. Use ./scripts/run-tracekit-auth.sh status to check, then guide the user through the TraceKit email verification flow with register and verify. That flow signs the user into an existing account for that email or creates the account automatically, then saves the returned credentials for the rest of the setup.

This is the entry point skill for TraceKit APM. It detects the user's technology stack, asks which features they want, and routes to the appropriate SDK and feature skills for setup.

When To Use

Use this skill when the user asks to:

  • Add TraceKit to their project (without specifying a language/framework)
  • Set up APM, observability, or error tracking (general request)
  • Get started with TraceKit
  • Choose the right TraceKit SDK for their project

Step 1: Detection

Before stack detection, ensure TraceKit auth exists for production setup flows. If the user is not connected yet, apply the tracekit-auth skill first. Do not redirect them to a separate signup or API-key-generation flow. The auth skill should handle the email verification flow, connect the existing account for that email or create it automatically, save the production credentials, and then continue directly with SDK setup.

Scan the project to determine the technology stack:

  1. Check for package.json -- Node.js/JavaScript ecosystem
    • next in dependencies => Use tracekit-nextjs-sdk skill
    • nuxt in dependencies => Use tracekit-nuxt-sdk skill
    • @angular/core in dependencies => Use tracekit-angular-sdk skill
    • react in dependencies (without next) => Use tracekit-react-sdk skill
    • vue in dependencies (without nuxt) => Use tracekit-vue-sdk skill
    • express, fastify, or @nestjs/core => Use tracekit-node-sdk skill
    • None of the above => Use tracekit-browser-sdk skill (vanilla JS/TS)
  2. Check for go.mod -- Use tracekit-go-sdk skill
  3. Check for requirements.txt or pyproject.toml -- Use tracekit-python-sdk skill
  4. Check for composer.json -- PHP ecosystem
    • laravel/framework in require => Use tracekit-laravel-sdk skill
    • Otherwise => Use tracekit-php-sdk skill
  5. Check for pom.xml or build.gradle -- Use tracekit-java-sdk skill
  6. Check for *.csproj or *.sln -- Use tracekit-dotnet-sdk skill
  7. Check for Gemfile -- Use tracekit-ruby-sdk skill

Step 2: Ask Which Features to Set Up

After detecting the stack, ask the user which TraceKit features they want to enable. Do not assume - present the options and let them choose. Users may want just error tracking, or the full suite.

Present the following feature menu:

Which TraceKit features would you like to set up? (you can pick multiple)

  1. Error Tracking - Automatic error capture with stack traces and context (included in base SDK setup)
  2. Distributed Tracing - Request tracing across services with performance monitoring (included in base SDK setup)
  3. Code Monitoring - Live breakpoints and snapshots for production debugging without redeploying (included in base SDK setup, enable with enable_code_monitoring=True)
  4. Custom Metrics - Track business KPIs, counters, gauges, and histograms alongside traces
  5. Session Replay (frontend only) - Record and replay user sessions linked to traces
  6. Source Maps (frontend only) - Readable stack traces from minified JavaScript
  7. Release Tracking - Monitor crash-free rates, deploy health, and regressions per release
  8. Alerts - Set up notifications for errors, performance degradation, and availability
  9. All of the above - Full observability suite

Default behavior: If the user says "just set it up" or doesn't specify, set up options 1-3 (Error Tracking, Distributed Tracing, and Code Monitoring) via the base SDK skill. Mention the other features are available and can be added later.

Step 3: Route to Skills

Based on the detected stack and selected features, apply skills in this order:

1. Base SDK Skill (always apply first)

Route to the SDK skill matching the detected stack. This sets up error tracking, distributed tracing, and code monitoring.

Backend SDKs

  • Go - tracekit-go-sdk skill (Gin, Echo, net/http)
  • Node.js - tracekit-node-sdk skill (Express, Fastify, NestJS)
  • Python - tracekit-python-sdk skill (Django, Flask, FastAPI)
  • PHP - tracekit-php-sdk skill
  • Laravel - tracekit-laravel-sdk skill
  • Java - tracekit-java-sdk skill (Spring Boot, Micronaut)
  • .NET - tracekit-dotnet-sdk skill (ASP.NET Core)
  • Ruby - tracekit-ruby-sdk skill (Rails, Sinatra)

Frontend SDKs

  • Browser (vanilla JS/TS) - tracekit-browser-sdk skill
  • React - tracekit-react-sdk skill (ErrorBoundary, React Router breadcrumbs)
  • Vue - tracekit-vue-sdk skill (Vue plugin, Vue Router breadcrumbs)
  • Angular - tracekit-angular-sdk skill (NgModule/standalone, DI ErrorHandler)
  • Next.js - tracekit-nextjs-sdk skill (multi-runtime, App Router/Pages Router)
  • Nuxt - tracekit-nuxt-sdk skill (Nuxt module, defineNuxtPlugin)

2. Feature Skills (apply based on user selection)

After the base SDK is set up, apply additional feature skills as selected:

  • Session Replay - tracekit-session-replay skill (frontend projects only)
  • Source Maps - tracekit-source-maps skill (frontend projects only)
  • Release Tracking - tracekit-releases skill
  • Alerts - tracekit-alerts skill
  • Custom Metrics - tracekit-custom-metrics skill
  • Distributed Tracing (multi-service) - tracekit-distributed-tracing skill (when connecting frontend + backend or multiple services)
  • Code Monitoring (advanced) - tracekit-code-monitoring skill (for programmatic snapshots beyond the base setup)

References

  • TraceKit docs root: https://app.tracekit.dev/docs
  • Dashboard: https://app.tracekit.dev
  • Quick start: https://app.tracekit.dev/docs/quickstart
Related skills

More from tracekit-dev/tracekit-for-ai

Installs
6
GitHub Stars
1
First Seen
Mar 9, 2026