Draft outreach, then decide if each lead is worth sending to
They sketched an agent workflow that rewrites sales outreach and gives a final send, wait, or reject recommendation.
- What they did
- The post outlined a system the person wanted to build, rather than a tool they had already finished. They said their manual cold outreach was already clearing a reply rate above 10%, and they hoped to automate the process with a scraping setup split between a CLI and a browser extension using native messaging. Their plan used three agent stages: one to produce an initial pitch, one to play the prospect and keep revising it, one to confirm facts and layout, and one last reviewer to choose whether the lead should move ahead, be held back, or be dropped.
- How it went
- No implementation was shared, so there was nothing to measure from the proposed system. The only performance number given was the existing manual reply rate of over 10%, and the discussion did not include any automation outcomes.
- Tips
- Keep the final go/no-go step separate from the writing loop so low-value prospects can be filtered out before anything is sent. If cost matters, set a firm per-outreach ceiling and require a human check before delivery.
Try it yourself with Claude
I want help with outreach for [prospect]. Use these inputs: [prospect background], [my current draft], and [facts to include]. First, produce a cleaner version, then make it more tailored to the prospect, then tighten the request. After that, check the draft for correctness, formatting, and tone, and recommend one of three outcomes: SEND, DEFER, or DISQUALIFY. Give your reasoned recommendation and the final text, but pause for my approval before anything is sent, submitted, or otherwise delivered.
Fill in the [brackets]. Agents can make mistakes, so review anything before it is sent, paid or deleted.
Get 3 agent uses like this every morning
With the day's top AI news. Free, and one click to unsubscribe.
Related agent uses
Automate renter replies and tour booking
An embedded leasing agent handled questions, tour scheduling, and follow-ups for multifamily listings.
Draft ticket replies and send daily support summaries
Use an agent to answer repeat support emails and compile short daily ticket digests for each teammate.
Set up daily social posts on autopilot
A CreaoAI agent drafts, queues, and monitors a business’s social posts every day without extra work after setup.
Set up a support agent that routes payment questions
They built a customer support agent that answers product and pricing questions, then passes payment proof and edge cases to staff.
Create a multi-source company brief automatically
This agent turns a company name into a structured research brief using business data, web signals, and review input.
Build a lead-gen pipeline for weak local websites
A user set up several agents to find neglected local businesses and draft tailored outreach for web services.