critical-loop

name	critical-loop
description	Create a critical-thinking loop for the current work.
user-invocable	true

Critical Loop Skill

Purpose

Scaffold, validate, review, and apply/test the current user request. This skill orchestrates the entire agent calling lifecycle: calling #agent dba -> debug -> performance-engineer -> quality-engineer -> refactoring-expert -> critical-thinking to iteratively refine the solution until it meets the user's needs.

When to Use This Skill

• User asks to "use critical-thinking loop", "run critical review loop", "enter critical loop", "usual thinking loop", or similar phrases indicating they want a comprehensive, iterative approach to a complex task
• Task involves multiple steps that require validation and expert review
• Any task that involves evolving the GeoKrety website PostgreSQL schema
• When exhaustive regression tests are needed alongside a schema change

Expert Review Loop

Invoke the following agent loop. Maximum 5 full rounds; if consensus is not reached by then, record remaining concerns in ${workspace}/99-OPEN-QUESTIONS.md and proceed. The human user has final authority on any unresolved concern. At the end of each round, check the updated files again as the user may have added TODO comments in the updated files to guide the next round of review, you must read those comments and verify they are addressed in the next round. Also re-reader this SKILL.md documentation after each round to ensure all best practices are followed as there may have been live updates to the documentation.

General Step-by-Step Workflow

1. Initial Analysis

   • Read the user's request carefully
   • Identify the main goal and any specific requirements
   • Determine which experts (DBA, Debug, Performance Engineer, Quality Engineer, Refactoring Expert) need to be involved, generally all of them for database-related tasks

2. DBA Phase

   • Generate SQL code for schema changes, migrations, or backfill scripts
   • Ensure code follows best practices for PostgreSQL and GeoKrety's architecture
   • Provide explanations for each change and how it fits into the overall goal

3. Debug Phase

   • Review the generated SQL for potential issues, edge cases, and unintended consequences
   • Simulate the execution of the SQL to identify any logical errors or performance bottlenecks
   • Refine the SQL based on findings

4. Performance Engineer Phase

   • Analyze the SQL for performance implications, especially on large datasets
   • Suggest optimizations such as indexing strategies, batch sizes, or query restructuring
   • Provide benchmarks or estimates for execution times

5. Quality Engineer Phase

   • Design comprehensive tests to validate the correctness of the SQL changes
   • Ensure tests cover edge cases and potential failure points
   • Provide instructions for running the tests and interpreting results

6. Refactoring Expert Phase

   • Review the SQL for maintainability and readability
   • Suggest improvements to make the code cleaner and easier to understand
   • Ensure that the code adheres to any style guides or conventions used in the GeoKrety codebase

7. Check for User Comments

   • After each round, check if the user has added any comments or TODO notes in the updates files after the previous rounds of review.
   • If yes, verify that the concerns raised in those comments are addressed in the next round of review.

8. Critical Thinking Phase

   • Review the entire solution holistically, ensuring it meets the user's original requirements and is robust against potential issues
   • Consider any edge cases or scenarios that may not have been covered in previous phases
   • Provide a final assessment of the solution's readiness for production deployment

9. Final Review

   • Pass unresolved concerns back through agents up to the 5-round cap.
   • If any concern remains, file it in ${workspace}/99-OPEN-QUESTIONS.md, add a cross-reference comment in the relevant file, and proceed
   • open questions do not block the full review process or database migration apply but must be reviewed before the next production deployment.