Guardrail Rules

Guardrail Check Script

Run scripts/guardrail-check.py <outline.md> <speaker-profile.json> to compute guardrail checks. The script handles the three-outcome PASS/WARN/FAIL logic:

• [FAIL] — value exceeds the limit
• [WARN] — value is under the limit but within 5 percentage points of it
• [PASS] — value is under the limit by more than 5 percentage points

The script checks: slide budget, Act 1 ratio, closing completeness, cut lines, data attribution, and profanity. Add the remaining checks (branding, time-sensitive, anti-patterns, illustration coverage, pattern strategy) to the report manually.

Antipattern Tags

Every antipattern flag MUST be tagged as one of:

• [RECURRING] — matches a pattern in the speaker's antipattern_frequency or recurring_issues profile history. The speaker has done this before.
• [CONTEXTUAL] — detected in the current outline but NOT in the speaker's historical profile. First-time issue for this talk.

Never use generic unlabeled antipattern warnings.

4-Tier Pattern Strategy

When recommending presentation patterns, organize into exactly four tiers:

1. Signature — patterns the speaker already uses consistently (from signature_patterns)
2. Contextual — patterns the speaker uses sometimes, worth considering here (from contextual_patterns)
3. New to You — patterns the speaker has never used but would fit this talk
4. Shake It Up — exactly 1-2 wild card patterns for experimentation. Never 0, never 3+.