wf-case-study-pipeline

Turn a customer conversation into a published case study and distribute it. Use when the user wants to create a case study from meeting notes or customer data and then promote it. Also trigger when th

SKILL.md

GitHub

Case Study Pipeline

End-to-end workflow: pull meeting notes, draft the case study, clean it up, then push it across social and email channels.

Stage 1: Gather Source Material

Ask the user for the source:

  • Granola meeting notes → Load and run the granola skill to pull notes
  • Raw notes or transcript → User pastes or provides a file
  • Existing draft → User points to a file to refine

Extract: customer name (or anonymous label), problem, solution, results, quotes.

Gate: Confirm the key facts with the user before drafting. Flag anything that needs [VERIFY].

Stage 2: Draft the Case Study

Load and run the case-studies skill for the full drafting workflow:

  • Branded or anonymous (ask if not specified)
  • Problem → Solution → Results structure
  • Pull metrics and quotes from source material
  • Include proof points from brain files

Gate: User approves the draft.

Stage 3: Quality Pass

Load and run the tagore skill on the approved draft. Fix the AI patterns and voice issues it flags.

Gate: Show the Tagore score. Get approval before distributing.

Stage 4: Social Distribution

Load and run the social-content skill. Using the case study as source, generate:

  • 1-2 LinkedIn posts highlighting key results
  • 1-2 Twitter/X posts with compelling stats or quotes
  • Any other platforms the user specifies

Pass forward: customer name (if branded), key metrics, the core narrative.

Gate: User approves social posts.

Stage 5: Email Integration (Optional)

Ask: "Want to add this case study to an email sequence or send it standalone?"

If yes, load and run the email-sequence skill to:

  • Draft a case study announcement email
  • Suggest placement in existing nurture sequences (e.g., bottom-of-funnel proof)

If no, skip to summary.

Summary

## Case Study Pipeline Complete

### Case Study
- [title] — [branded/anonymous] → [path]
- Customer: [name or descriptor]
- Key result: [headline metric]

### Social Posts
- LinkedIn: [count] posts → [paths]
- Twitter/X: [count] posts → [paths]

### Email
- [what was created, or "Skipped"]

### Next Steps
- [ ] Get customer approval (if branded)
- [ ] Publish case study
- [ ] Schedule social posts
- [ ] Add to website case studies section
- [ ] Reference in sales enablement materials

Output Structure

All files go into a single project folder:

marketing/case-studies/[customer-slug]/
  case-study.md
  social-linkedin-1.md
  social-linkedin-2.md
  social-twitter-1.md
  email-announcement.md         (if email stage run)
  email-nurture-insert.md       (if email stage run)

Use the customer name (or anonymous descriptor) as the slug. Confirm with the user before creating files.

Rules

  • Each stage uses the full component skill
  • Always gate between stages
  • Never publish customer names or metrics without noting they need customer approval
  • All claims must be verifiable from the source material — do not embellish results
  • If the user wants to skip a stage, skip it
  • All output goes into the project folder, not scattered across marketing/