| name | personalization-writer |
| description | Generate personalized first lines, email copy, or LinkedIn messages based on prospect data and copy frameworks. |
Personalization Writer
Task
Generate personalized outreach messages for a list of prospects. Uses the copy frameworks defined in your CLAUDE.md and adapts each message based on the prospect's specific signals and context.
Input Requirements
- An enriched prospect list (CSV or JSON) with at minimum:
contact_name, company_name, contact_title, signal, personalization_note
- Message type:
first_line | full_email | linkedin_connection | linkedin_dm | follow_up
- Messaging angle:
signal_based | problem_led | social_proof | content_based
- Sequence position (for follow-ups):
follow_up_1 | follow_up_2 | breakup
Instructions
- Read the prospect list and the copy frameworks section of your CLAUDE.md
- For each prospect:
- Identify their primary signal (hiring, funding, tech adoption, content, etc.)
- Select the matching message template from copy frameworks
- Research additional context if needed (recent LinkedIn posts, company news)
- Write the personalized message following tone rules and length guidelines
- Quality checks per message:
- Contains at least one detail specific to this prospect (not reusable for anyone else)
- Under the word limit for the message type
- CTA is a soft ask (never "book a demo")
- No buzzwords, no exclamation marks, no emojis
- Sounds like a human wrote it, not a template
- Flag for review: any prospects where personalization context is thin - don't force a bad personalization
Output Format
CSV with columns:
contact_name, contact_email, company_name, message_type, subject_line, message_body, personalization_source, confidence_score
confidence_score: high (strong signal + specific context), medium (decent context), low (thin personalization - review before sending)- Save to:
./data/outreach/[DATE]_[MESSAGE_TYPE]_personalized.csv
Example Usage
/personalization-writer
Input: ./data/enriched/saas_companies_march_enriched.csv
Message type: full_email
Angle: signal_based
Output:
contact_name: Alex Chen
company_name: DataFlow Inc
subject_line: Your 3 new data engineer roles
message_body: Alex, saw DataFlow is hiring 3 data engineers - scaling
the pipeline team. Most data teams at your stage hit observability
issues around month 3. We helped Acme cut pipeline debugging by 70%.
Worth a quick look?
personalization_source: LinkedIn job posting (3 data engineer roles, posted 5 days ago)
confidence_score: high
