Sales outreachCAN I VIBECODE
OutFriend?
BUILD FIRST
- One specific customer type and one primary use case
- A vertical slice that completes the main job without manual workarounds
- Visible limits, review steps, and a fallback when an external dependency fails
LEAVE FOR LATER
- Broad third-party integrations
- High-scale collaboration, syncing, and permissions
- Automation that can cause harm without human review
RESEARCH NOTES
What this tool actually does.
OutFriend is a cold-email outreach tool whose core promise is "cold email that sounds like you." It learns a user's writing voice from emails they have actually sent, then generates outreach against a target brief the user defines: industries, company size, and job titles. The workflow is: (1) define goals/brief; (2) upload leads via CSV — unverified, bounced, and opted-out contacts are filtered before anything is queued; (3) choose a rollout plan — send days, sending hours, and a daily cap, with every draft reviewable and editable before sending; (4) the system sends emails in the user's voice, auto-schedules follow-ups that halt the moment a prospect replies, and keeps schedules, drafts, sends, and replies in one place. Marketing tagline: "Your next 10 meetings are one CSV away." No verified pricing, detailed technical documentation, or secondary-source reviews were available in the se
RESEARCHED BUILD PROMPT
Give this to your coding agent.
Build 'VoiceOutreach', a focused replacement for the OutFriend-style workflow: cold email that is written in the user's own voice. Core loop: (1) Voice Profile — the user uploads 5–10 emails they have genuinely sent; your app extracts tone, phrasing, sentence length, and vocabulary to create a style fingerprint. (2) Target Brief — the user specifies industries, company sizes, and job titles; every generated email is constrained by this brief. (3) Lead Import — the user uploads a CSV of leads; the app verifies each email, drops invalid/bounced/opted-out addresses, and shows a filtered review list before sending. (4) Rollout Plan — the user picks send days, sending hours, and a daily cap, and can read/edit every generated draft before approval. (5) Send & Follow-up — approved emails are sent on schedule; follow-ups are auto-generated in the same voice and stop immediately when a reply is detected; the full thread appears in the dashboard.
User-facing requirements: a dashboard with sections for Voice Profile, Target Brief, Leads, Draft Review, Schedule, and Replies; inline draft editing; CSV upload with per-row validation status; per-campaign send window and daily cap controls; reply detection and auto-stop of follow-ups; a simple activity log. Technical requirements: Next.js/React front end; Node or Python API; Postgres for campaigns/leads/threads; BullMQ/Redis for scheduled sends; transactional email via Resend or SendGrid with bounce webhooks; email verification via ZeroBounce or similar; style transfer via OpenAI/Anthropic few-shot prompting using the user's own email samples as exemplars; reply detection via IMAP polling or provider webhooks. Implement voice learning as few-shot prompt assembly rather than fine-tuning to keep setup simple. Do NOT copy OutFriend's brand, name, copy, or UI layout — create an original clean interface. Explicitly out of scope: multi-channel outreach (LinkedIn/WhatsApp/calls), lead generation/discovery, A/B testing, team collaboration/roles, billing, Chrome extensions, and advanced analytics. Deliver a README covering setup steps (env vars, DB migration, queue start, webhook endpoints) and important limitations (voice quality depends on sample quality, verification catches only some bounces, sending limits are enforced locally not by providers, and the tool must support unsubscribe/opt-out compliance).