
5 Claude Code prompts I run every week against my outreach database
Copy-paste agent prompts for client briefings, workspace lead status, deliverability, segment performance, reply copy insights, and lead context - powered by Outreach Magic local SQLite.
I sync Smartlead, Instantly, and Heyreach into one SQLite file via Outreach Magic. Here are six prompts I run weekly - each one replaces a dashboard session or CSV export.
1. Client check-in briefings
Prompt: "Prepare next Tuesday's client agenda for Acme Corp. Include the week's campaign performance, standout replies, deliverability issues, and the three things I should bring up first on the call."
Expected output: A structured agenda with metrics, quoted replies, bounce flags, and prioritized talking points.
What's underneath: Queries campaigns, replies, bounces, and lead_timeline scoped to the client.
2. Workspace lead status plans
Prompt: "Build me a workspace lead status plan from healthiest to weakest across all my Smartlead sub-accounts. Tier them as top picks, backups, and exclusions, with a one-line reason for each."
Expected output: Three tiers of senders/domains with reasons, ready to paste into Smartlead.
What's underneath: Reads senders, bounces, and domain health signals across accounts.
3. Lead history insights
Prompt: "Why are we bouncing this week? Break it down by sender domain, day, and reseller. Tell me what to do about it."
Expected output: A breakdown table plus recommended actions (pause sender, rotate domain, etc.).
What's underneath: Aggregates bounces by domain, day, and reseller metadata.
4. Segment performance
Prompt: "Show me reply rate by seniority for the Q4 outbound campaign, split by country. Then tell me which segment to double down on and which to kill."
Expected output: Segment table with reply rates and a clear kill/double-down recommendation.
What's underneath: Joins lead_attributes, replies, and campaigns.
5. Reply copy insights
Prompt: "Which subject lines earned replies in the past 14 days across all campaigns? Group them by theme."
Expected output: Themed clusters of winning subject lines with reply counts.
What's underneath: Pulls reply content and subject metadata from replies joined to events.
6. Lead timeline + AI context (bonus)
Prompt: "Show everything we know about Sarah from Acme - email touches, LinkedIn touches, enrichment data. Then draft a follow-up that picks up where the last thread left off."
Expected output: Full timeline summary plus a contextual follow-up draft.
What's underneath: Reads lead_timeline, events, and lead_attributes for one lead.
---
Each prompt assumes Outreach Magic is installed and webhooks are flowing. Get started → · Full prompt reference →