cd /news/artificial-intelligence/claude-fable-5-prompt · home topics artificial-intelligence article
[ARTICLE · art-25572] src=gist.github.com pub= topic=artificial-intelligence verified=true sentiment=· neutral

Claude Fable 5 Prompt

A developer has created a detailed prompt for Claude Fable 5 that instructs the AI to act as a Principal Software Architect and Performance Engineer, performing a comprehensive performance audit and refactor of an entire codebase. The prompt requires the AI to identify all performance bottlenecks—including N+1 queries, inefficient SQL, caching issues, and memory problems—and produce a PERFORMANCE_AUDIT.md and PERFORMANCE_PLAN.md before implementing any changes. The approach prioritizes optimizations by ROI, with high-impact, low-risk fixes first, while preserving all existing functionality and maintaining code readability.

read2 min publishedJun 9, 2026

Act as a Principal Software Architect and Performance Engineer.

Your task is to perform a comprehensive performance audit and refactor of this entire codebase.

Before making any changes:

Thoroughly analyze the architecture, code structure, dependencies, database access patterns, caching strategy, queues, API endpoints, background jobs, asset , and deployment configuration.

Identify all performance bottlenecks, including but not limited to:

  • N+1 database queries

  • Missing or ineffective indexes

  • Inefficient SQL queries

  • Excessive memory allocations

  • Unnecessary object creation

  • Duplicate computations

  • Blocking I/O operations

  • Excessive network requests

  • Large payloads

  • Inefficient loops and algorithms

  • Slow API endpoints

  • Cache misses and poor cache utilization

  • Queue bottlenecks

  • Expensive model relationships

  • Over-fetching data

  • Asset and frontend performance issues

Produce a detailed PERFORMANCE_AUDIT.md containing:

  • Identified bottlenecks

  • Estimated impact

  • Root cause analysis

  • Recommended fixes

  • Risk assessment

  • Expected performance gains

Prioritize optimizations by ROI:

  • High impact / low risk first
  • High impact / medium risk second
  • Everything else afterward

When implementing improvements:

  • Preserve all existing functionality.
  • Do not introduce breaking changes unless absolutely necessary.
  • Maintain readability and maintainability.
  • Follow existing project conventions.
  • Add or update tests when required.
  • Avoid premature optimization.

For every significant change, explain:

  • What was changed
  • Why it was changed
  • Expected impact on:
  • Response time
  • Throughput
  • Memory usage
  • Database load
  • Infrastructure cost

Specific areas to investigate deeply:

  • Query plans

  • Index usage

  • Composite indexes

  • Eager opportunities

  • Pagination strategy

  • Aggregate queries

  • Bulk operations

  • Service architecture

  • Object lifecycle

  • Dependency resolution

  • Serialization costs

  • Validation overhead

  • Event/listener performance

  • Query caching

  • Response caching

  • Redis utilization

  • Cache invalidation strategy

  • Long-running jobs

  • Retry policies

  • Batch processing opportunities

  • Queue throughput

  • Payload size reduction

  • Serialization optimization

  • Endpoint response times

  • HTTP caching opportunities

  • Bundle size

  • Lazy

  • Asset optimization

  • Render performance

  • Runtime configuration

  • Connection pooling

  • Worker configuration

  • Container optimization

Execution Strategy:

  • Audit
  • Create plan
  • Present findings
- Implement highest-impact improvements
- Re-measure
  • Continue iteratively

Do not start refactoring immediately.

First create PERFORMANCE_AUDIT.md and PERFORMANCE_PLAN.md, then wait for approval before making changes.

── more in #artificial-intelligence 4 stories · sorted by recency
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
Live at https://your-agent.zahid.host
Get free account → Pricing
from €0/mo · no card required
LIVE [news/claude-fable-5-promp…] indexed:0 read:2min 2026-06-09 ·