confidence-scorelisted

# Confidence Score Not all recommendations are equal. Know which ones to trust and which ones to question. ## When to Activate - User says `/confidence` or `/confidence status` - Auto-applied: every skill output should include confidence tagging - User asks "how sure are you about this?" ## Commands ### `/confidence` — System-wide confidence report ### `/confidence {skill}` — Confidence level for a specific skill ### `/confidence calibrate` — Recalibrate based on prediction vs actual ## Confidence Levels | Level | Symbol | Basis | Trust Level | |-------|--------|-------|-------------| | Very High | ★★★★★ | 30+ data points + validated by A/B test | Act on this | | High | ★★★★ | 15-29 data points from performance-log | Strongly consider | | Medium | ★★★ | 5-14 data points or cross-referenced estimates | Use with judgment | | Low | ★★ | < 5 data points or single-source estimate | Treat as hypothesis | | Guess | ★ | No data, pure heuristic/general knowledge | Test before trusting | ## Workflow ### System-Wide Report ``` ============================================ System Confidence Report ============================================ Data foundation: content-history entries: {N} Entries with performance data: {N}/{total} ({percentage}%) A/B tests completed: {N} DNA last updated: {date} Skill confidence levels: HIGH CONFIDENCE (★★★★+): content-dna: ★★★★ — based on {N} entries with performance best-time: ★★★★ — confirmed by {N} posts across time slots sm