Copilot CLI Turbo Prompt Writing Voice

# Copilot CLI Turbo Prompt - Writing Voice

Paste this into Copilot CLI for a faster run.

---

You are operating my Writing Voice project.

Goal:
Learn my writing voice from my authored email/docs/decks and produce drafts that sound like me.

Execution rules:
- Be proactive and execution-first.
- Do not stop at planning.
- Ask only if blocked by access/permissions.
- Prefer updating project files over long chat responses.
- Never invent facts; mark uncertainty clearly.

Data source:
- Use WorkIQ for M365 context (sent mail, authored docs, authored decks).

Project files:
- Plan: Writing Voice/Writing-Voice-Project-Plan.md
- Query list: Writing Voice/Query Packs.md
- Inventory output: Writing Voice/data/raw/inventory-summary.md
- Gold set: Writing Voice/data/processed/gold-set-index.md
- Voice profile: Writing Voice/prompts/voice-profile.md
- Weekly eval: Writing Voice/evals/weekly-report.md

Run now in this order:

1. Confirm WorkIQ is available.
2. Execute Query Packs A-D from Writing Voice/Query Packs.md.
3. After each pack, append structured dated output to Writing Voice/data/raw/inventory-summary.md.
4. Curate corpus: remove low-signal items (one-liners, boilerplate-only, forwarded-only).
5. Build balanced gold set and populate Writing Voice/data/processed/gold-set-index.md with tags:
   - artifact_type, intent, audience, tone, length, date
6. Update Writing Voice/prompts/voice-profile.md with:
   - tone by audience
   - structure templates
   - preferred phrases/transitions
   - avoid list
   - rewrite rules
7. Create/refine drafting prompts for:
   - email reply, email outreach, status update, one-pager
8. Start draft loop for requests:
   - retrieve 3-5 similar gold examples
   - generate v1
   - style alignment pass
   - clarity pass
   - return short/standard/executive versions
9. Update Writing Voice/evals/weekly-report.md with scores:
   - voice match, clarity, actionability, accuracy (1-5)
   - edit taxonomy (tone, structure, vocabulary, correctness)

MVP targets:
- 300 curated samples reviewed
- 100 high-fidelity gold samples indexed
- average email voice match >= 4/5
- median manual edit effort down >= 40%

Incremental sync — "sync writing voice":
Run Writing Voice/Sync Writing Voice.md workflow:
1. Read last_synced from inventory-summary.md Sync Metadata.
2. Pull new artifacts via WorkIQ since that date.
3. Append to inventory, curate into gold set, re-analyze style, update metrics.
4. Bump last_synced. Print summary.

Output format each cycle:
- What was updated (files)
- Key decisions made
- Next concrete action