repo-assessing-criticality-confidence

What problem does it solve?

Checker and fixer agents often use inconsistent criteria to prioritize audit findings, leading to critical issues being missed, fixes being applied in the wrong order, and audit reports that lack uniform structure and context for stakeholders.

Core Features & Use Cases

• Two-dimension classification system: Separate criticality (urgency/importance of the issue) and confidence (certainty the issue is real and fixable) to eliminate ambiguity in prioritization.
• P0-P4 priority matrix: Clear, rules-based execution order for fixers to address blocker issues first, followed by urgent, standard, and low-priority items.
• Dual-label pattern: Combine factual verification status (e.g., [Error], [BROKEN]) with criticality level to provide full context for each finding.
• Standardized report templates: Pre-defined formats for checker audit reports and fixer fix reports to ensure consistency across all agent outputs.
• Domain-specific guidance: Concrete examples for governance rule checks, documentation audits, and crud application content validation to speed up implementation. Use case: A docs-link-checker agent can flag a broken internal link as CRITICAL, and a docs-fixer agent can assess HIGH confidence to auto-fix the link immediately, or MEDIUM confidence to flag it for manual review if the correct target is ambiguous.

Quick Start

Use this skill to classify all audit findings by criticality and confidence, then apply fixes in strict P0-to-P4 priority order to resolve the most urgent issues first.