Restore Read the Room writing principles

[letsrokk]

Summary

Restore explicit writing guidance lost during prompt compression, organizing Make It Make Sense around Zinsser's Clarity, Simplicity, Brevity, and Humanity.

• Incorporate Google documentation practices and selected ASD-STE100 ambiguity checks while preserving natural voice, necessary detail, and evidence limits.
• Keep startup guidance focused on universal decisions; load documentation details and source rationale on demand. The measured startup payload is 5.9 KB, about 795 words.
• Add fixed writing scenarios for future policy comparisons and verify that the source reference stays out of startup context.

Verification

• npm test
• npm run validate
• Read the Room: node scripts/test.js and node scripts/validate.js
• Skill frontmatter validator and git diff --check
• Four qualitative writing scenarios showed no material regression. The comparison used the same agent context and does not establish measured improvement across models or sessions.

Review considerations

The startup policy is longer to preserve meaningful distinctions. Supporting detail remains on demand, and the skill stays below the repository's 7,168-byte limit. Hook tests verify loading, not prose quality; the plugin does not claim STE compliance.