Implementation Confidence Check FAQs

Question 1

Does it check for existing code duplicates?

Accepted Answer

Yes, it uses Grep and Glob searches to scan your codebase for existing functionality, ensuring that the AI doesn't reinvent solutions that already exist in your project.

Question 2

How does this skill save tokens?

Accepted Answer

By spending 100-200 tokens on a preliminary check, you avoid spending 5,000-50,000 tokens on large-scale implementations that might be architecturally flawed or redundant.

Question 3

What happens if the confidence score is low?

Accepted Answer

If the score is between 70% and 90%, the skill presents alternatives and asks clarifying questions. If it falls below 70%, it completely stops implementation to request more context.

Question 4

What is the 90% confidence threshold?

Accepted Answer

It is a mandatory scoring requirement where the skill evaluates five criteria; if the combined score is below 0.90, the AI is instructed to stop or ask for more context rather than proceeding with potentially incorrect implementation.

Question 5

How does it verify official documentation?

Accepted Answer

The skill integrates with Context7 and WebFetch MCPs to review the latest official documentation and API specifications, ensuring the proposed solution isn't based on outdated assumptions.

Implementation Confidence Check

关于

主要功能

使用场景

Implementation Confidence Check

关于

主要功能

使用场景