feat: improve skill scores for 5 codeywood skills

[rohan-tessl]

Hey @kaigani 👋

I ran your skills through tessl skill review at work and found some targeted improvements.

Here's the full before/after:

Skill	Before	After	Change
story-intake	10%	90%	+80%
character-architect	10%	79%	+69%
canon-database-manager	10%	77%	+67%
prop-reference-generator	5%	74%	+69%
visual-continuity-validator	5%	74%	+69%

This PR is intentionally scoped to 5 skills to keep it reviewable — more skills can be improved in follow-ups or via automated review on future PRs.

Changes summary

All 5 skills:

• Added YAML frontmatter with name and description fields (required for validation and discoverability)
• Description includes a "Use when..." clause for proper trigger matching
• Restructured sections from generic Purpose/Trigger/Inputs to cleaner Inputs/Outputs/Process format

story-intake:

• Tightened interview instructions with explicit constraints (ask one at a time, max 10)
• Added downstream pipeline context in Notes (visual keywords feed reference generators)
• Quality gate reformatted with clear pass/fail criteria and recovery path

character-architect:

• Consolidated character tiers into a scannable table
• Condensed psychology deep-dive into a numbered list format
• Preserved all 8 steps including voice differentiation test and contradiction check

canon-database-manager:

• Merged 10 steps into 7 without losing coverage (grouped factions + canon facts + continuity log)
• Added explicit error recovery guidance in update protocol
• Referenced external schema file for progressive disclosure

prop-reference-generator:

• Added priority table (signature/plot/environment) for quick reference
• Included inline prompt template example with substitution variables
• Reduced redundancy between categories section and process steps

visual-continuity-validator:

• Added tolerance table (within scene vs. across scenes vs. across episode)
• Consolidated drift types into a clear 3-type classification
• Moved end-of-clip video review into a compact subsection

What does this PR do?

Adds YAML frontmatter and improves structure for 5 skills that were scoring 5–10% on tessl skill review due to missing metadata. After optimization, scores range from 74–90%.

Why is this change needed?

Without frontmatter, the skills fail deterministic validation and the LLM evaluation is skipped entirely — meaning agents can't properly discover or score them. These changes make the skills machine-readable while preserving all domain expertise.

How was this tested?

Each skill was evaluated with tessl skill review before and after changes. No behavioral changes to the skills themselves — only structural improvements for discoverability and validation.

Checklist

• Follows existing skill structure
• Includes relevant templates (preserved existing references)
• Examples provided (inline where applicable)
• Documentation updated (frontmatter + descriptions)

---

Honest disclosure — I work at @tesslio where we build tooling around skills like these. Not a pitch - just saw room for improvement and wanted to contribute.

Want to self-improve your skills? Just point your agent (Claude Code, Codex, etc.) at this Tessl guide and ask it to optimize your skill. Ping me - @rohan-tessl - if you hit any snags.

Thanks in advance 🙏