Machine-learning Services Bug Hunt & Fix Protocol Assistant
A systematic debugging workflow for machine-learning services that finds root causes instead of symptoms. Built for platform teams working in machine-learning services, tuned for Claude. Fill in the placeholders, paste, and get a structured, ready-to-use result — with clear rules so the output stays specific to your situation instead of generic.
Preview
You are a code-quality reviewer for teams shipping machine-learning services.
Inputs you will receive:
Feature: {{FEATURE}}
Constraints: {{CONSTRAINTS}}
Codebase context: {{CODEBASE_CONTEXT}}
Stack: {{STACK}}
… [purchase to unlock the full framework]Screenshots
Example outputs
H1 (70%): stale cache key after tenant switch. H2 (20%): timezone in cron window. H3 (10%): …
Version history
- v1Initial releaseJul 10, 2026
Buyers get every future update free.
Buyers also bought
Paste a slow query and schema — get back an optimized version, index recommendations, and a plain-English explanation of why it was slow.
Safely refactor gnarly legacy code: characterization tests first, seams identification, strangler-fig migration plan. Based on real enterprise migration experience.
A disciplined build-a-feature workflow for REST APIs: plan first, tests first, small diffs. Built for agency dev shops working in REST APIs, tuned for Cursor. Fill in the placeholders, paste, and get a structured, ready-to-use result — with clear rules so the output stays specific to your situation instead of generic.
A safe refactor plan for machine-learning services: characterization tests, seams, and steps that keep CI green. Built for platform teams working in machine-learning services, tuned for Windsurf. Fill in the placeholders, paste, and get a structured, ready-to-use result — with clear rules so the output stays specific to your situation instead of generic.
More from James Castillo
A thorough review pass tuned for Python backends — correctness first, style last. Built for solo developers working in Python backends, tuned for Claude. Fill in the placeholders, paste, and get a structured, ready-to-use result — with clear rules so the output stays specific to your situation instead of generic.
A safe refactor plan for browser extensions: characterization tests, seams, and steps that keep CI green. Built for solo developers working in browser extensions, tuned for Claude. Fill in the placeholders, paste, and get a structured, ready-to-use result — with clear rules so the output stays specific to your situation instead of generic.
A structured way to make and document architecture calls for serverless functions. Built for technical founders working in serverless functions, tuned for Claude. Fill in the placeholders, paste, and get a structured, ready-to-use result — with clear rules so the output stays specific to your situation instead of generic.
Meaningful test coverage for GraphQL services — behavior tests, edge cases and fixtures that stay maintainable. Built for solo developers working in GraphQL services, tuned for Claude. Fill in the placeholders, paste, and get a structured, ready-to-use result — with clear rules so the output stays specific to your situation instead of generic.
Reviews
No reviews yet.