confidence-check

Execute a five-check confidence assessment with weighted scoring and threshold-driven decisions.

2|Updated Sep 30, 2025
One-click install
npx skills add https://github.com/krzemienski/shannon-framework --skill confidence-check-krzemienski
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: confidence-check
Source: https://github.com/krzemienski/shannon-framework/tree/main/skills/confidence-check
Command: npx skills add https://github.com/krzemienski/shannon-framework --skill confidence-check-krzemienski

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Provides a quantitative, 5-check confidence algorithm to prevent wrong-direction work and enforce objective validation before implementation.

Core Features & Use Cases

  • Implements a weighted 5-check algorithm: duplicate verification, architecture compliance, official docs, working OSS references, and root-cause identification.
  • Threshold-driven decision: ≥90% PROCEED, 70-89% CLARIFY, <70% STOP.
  • Anti-rationalization guidance to prevent authority or urgency from bypassing validation.

Quick Start

Example: /shannon:confidence-check "Implement Redis caching" --save

Frequently Asked Questions about confidence-check

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I validate a software change before implementation?

Validation checks duplicate work, architecture alignment, official documentation, working OSS references, and root causes using a weighted 5-check algorithm. Scores ≥90% proceed, 70-89% clarify concerns, below 70% stop and reassess to prevent wrong-direction effort.

What does a confidence assessment check for architecture compliance?

Architecture checks evaluate whether your proposed change aligns with existing system design patterns and constraints. This 25%-weighted check prevents decisions that conflict with established technical direction and increases implementation confidence.

Can I use this validation for any type of feature work?

Yes, the 5-check framework applies across software projects from baseline through refactor phases. It guides decision-making on any feature by requiring evidence across duplicates, architecture, docs, OSS patterns, and root-cause analysis before proceeding.

What prevents urgency or authority from bypassing validation?

Anti-rationalization guidance enforces the threshold-driven scoring system objectively. The algorithm requires all five checks to reach confidence thresholds—no override flags bypass the validation gates, ensuring consistent rigor regardless of pressure.

How are the five checks weighted in the confidence score?

Duplicate verification and architecture each count 25%, official docs 20%, working OSS references 15%, and root-cause analysis 15%. This weighting balances existing work awareness with design fit, documentation grounding, and proven patterns to produce a deterministic confidence score.

When should I use the clarify threshold instead of stopping?

Scores between 70-89% signal clarify mode: the change shows promise but gaps remain. Use this threshold when you need to address specific concerns through documentation, stakeholder review, or design iteration before committing to implementation.