Instruction file imported from fabioc-aloha/Alex_Plug_In (
.github/instructions/ai-writing-avoidance.instructions.md). Copyright stays with the author.
AI Writing Avoidance — Auto-Loaded Rules
Full vocabulary ban list, structural rules, tone corrections with examples → see ai-writing-avoidance skill.
Quick Audit Pattern
Before finalizing any document:
- Ctrl+F for banned vocabulary (delve, myriad, plethora, tapestry, beacon, landscape, realm, paradigm)
- Check first paragraph for preambles ("In this document, we will explore...")
- Check last paragraph for restating conclusions
- Count bullet lists (max 3 per page)
- Verify at least one specific example exists
- Confirm document has a point of view
Policy Document Rules
- Lead with what people must DO
- Use role names, not "stakeholders"
- Include real incident references where appropriate
- State consequences directly
- Keep paragraphs under 4 sentences
Red Flag Thresholds
| AI Tells Found | Action Required |
|---|---|
| 0-2 | Minor polish |
| 3-5 | Section rewrite |
| 6+ | Full document revision |
- Pair with
documentation-quality-assurancefor systematic reviews - Use before any external publication
- Apply to email templates and form letters
- Required for customer-facing policy documents