tkm:confidence

Tag agent outputs with EXTRACTED, INFERRED, or AMBIGUOUS confidence levels.

Updated May 13, 2026

Search Tags:#code review #root cause analysis #confidence #tagging #agent outputs

Ready to run

npx skills add https://github.com/khapn-2933/agentic-coding-hands-on --skill tkm-confidence

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: tkm:confidence
Source: https://github.com/khapn-2933/agentic-coding-hands-on/tree/main/.claude/skills/confidence
Command: npx skills add https://github.com/khapn-2933/agentic-coding-hands-on --skill tkm-confidence

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill provides a standardized confidence taxonomy for agent outputs, allowing for the tagging of findings as EXTRACTED, INFERRED, or AMBIGUOUS with 0-1 scores.

Core Features & Use Cases

Confidence Tagging: Assign confidence levels to agent outputs for findings.
Taxonomy Reference: A reference skill that other skills import for consistent confidence tagging.
Use Case: For generating reports, debugging root cause analysis, and architecture discovery.

Quick Start

Load the confidence skill and apply it to the agent's output.

Frequently Asked Questions about tkm:confidence

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

FAQPage Schema

How do I tag agent outputs with confidence levels for code review findings?▼

Tag agent outputs with confidence levels by assigning EXTRACTED, INFERRED, or AMBIGUOUS tags based on the presence or absence of evidence in source code, configuration, or data. This requires consistent tagging conventions across your analysis outputs.

What is confidence tagging in root cause analysis?▼

Confidence tagging in root cause analysis labels debugging findings as EXTRACTED, INFERRED, or AMBIGUOUS with 0-1 scores. This standardized taxonomy reflects whether conclusions are directly pulled from evidence or deduced through inference.

How do I apply confidence scores to architecture discovery insights?▼

Apply confidence scores to architecture insights by evaluating the evidence supporting each finding. Insights directly read from source code or configuration are tagged EXTRACTED, while deduced architectural patterns are tagged INFERRED.

Can I use this confidence taxonomy to standardize reports across different analysis tasks?▼

Yes, you can use this confidence taxonomy to standardize reports across different analysis tasks. It provides a reference skill that other skills import to ensure consistent confidence tagging for code reviews, debugging, and architecture discovery.

What is the difference between EXTRACTED and INFERRED confidence levels?▼

EXTRACTED confidence levels apply to findings directly pulled from evidence in source code or configuration, while INFERRED levels apply to conclusions deduced without direct evidence. AMBIGUOUS tags findings where evidence presence is unclear.

Do I need specific dependencies to implement confidence tagging on agent outputs?▼

No specific dependencies are required to implement confidence tagging on agent outputs. The skill operates independently using scripts and references to apply the EXTRACTED, INFERRED, and AMBIGUOUS taxonomy to your analysis results.