Debug a Production Error
Quickly analyze a production error with a structured Root Cause Analysis approach and an immediate action plan.
Paste in your AI
Paste this prompt in ChatGPT, Claude or Gemini and customize the variables in brackets.
Tu es un ingénieur de garde spécialisé dans le débogage d'incidents en production. J'ai une erreur critique qui affecte [NOMBRE_UTILISATEURS] utilisateurs depuis [DURÉE]. Voici les informations disponibles :
Stack trace :
[COLLER_LA_STACK_TRACE]
Logs applicatifs (dernières 50 lignes) :
[COLLER_LES_LOGS]
Contexte :
- Stack technique : [EX: Node.js 18, PostgreSQL 14, Redis 7]
- Dernier déploiement : [DATE_ET_HEURE]
- Changements récents : [DESCRIPTION_DES_CHANGEMENTS]
- Environnement : [PROD/STAGING]
Analyse cette situation en suivant une approche structurée :
- Identifie la cause racine probable (Root Cause Analysis)
- Liste 3 hypothèses par ordre de probabilité avec leur justification
- Propose des étapes de diagnostic immédiates pour confirmer la cause
- Fournis un plan de correction immédiate (hotfix) et une solution long terme
- Suggère des alertes ou métriques à ajouter pour prévenir ce type d'incident
Priorise la rapidité de résolution car l'incident est actif.
Personalize this prompt with Léa
Léa rewrites this prompt for your job and your exact goal — 3 quick questions.
Why this prompt works
<p>This prompt replicates the mental process of an experienced SRE engineer facing a production incident. By providing structured context (stack trace, logs, recent changes), the AI can formulate relevant hypotheses rather than generic answers.</p><p>Requesting 3 hypotheses ranked by probability forces the AI to reason rather than propose the first solution that comes to mind. This considerably increases diagnostic accuracy, especially for complex bugs related to inter-service interactions.</p><p>Adding the question about alerts to set up transforms a simple debugging session into a systemic improvement opportunity, which is the hallmark of a senior engineer thinking in terms of long-term reliability.</p>
Use Cases
Expected Output
A root cause analysis with prioritized hypotheses, diagnostic steps, immediate hotfix, and preventive recommendations.
Improve this prompt
Run this prompt through the Optimizer to strengthen its context, constraints and expected format.
Improve this prompt with the OptimizerComments
- LéaAI
Ajoutez `request_id` ou `trace_id` à votre prompt pour corréler les logs avec l’exécution, et demandez l’arbre des appels (span) via votre APM. Cela affine les hypothèses RCA et accélère le diagnostic.
📬 Get new prompts every week
Join our newsletter and never miss a prompt.
Go further
Similar Prompts
Implement CSP with nonces
Prevent XSS with strict CSP
Implement JWT authentication
Add JWT auth to an API
GitHub Copilot Prompt for Generating JavaScript Code
GitHub Copilot, the AI-powered programming assistant developed by GitHub and OpenAI, has transformed the way developers write JavaScript code. By leveraging billions of lines of open-source code, Copilot can generate complete functions, classes, modules, and even entire architectures from simple natural language descriptions. To get the most out of this tool, mastering the art of prompt engineering for code generation is essential. A well-structured prompt produces clean, performant JavaScript code that adheres to modern best practices (ES2024+, implicit typing, error handling, modularity). In this guide, we offer optimized prompts for GitHub Copilot that will help you generate professional-grade JavaScript code, whether you're developing a REST API, a frontend component, a utility script, or complex business logic. Each prompt is designed to provide sufficient context to Copilot while allowing the flexibility needed for relevant and project-adapted generation.
Optimize Slow SQL Queries
Analyze and optimize your slow SQL queries through execution plan analysis and precise index recommendations.