prospecting

Use when auditing a prospect’s website without backend access or credentials. Also use when an agency needs to generate both internal technical findings and a client-facing business impact report for a new prospect.

Model	Source
sonnet	pack: client-ops

Full Reference

prospecting

External-only audit pipeline for prospect websites. No backend access, no credentials, no client cooperation required — everything runs against public-facing data. Given a URL and optional business name, produces two deliverables: an internal technical brief (agency eyes only) and a client-facing business impact report (shareable with prospect). Both outputs are generated via the client-reports skill using the prospecting-internal and prospecting-client templates. The pipeline runs one sequential crawl step first, then dispatches six parallel analysis agents covering performance, Core Web Vitals, schema, local SEO, on-page SEO, and AI visibility.

Quick Reference

Item	Value
Input	audit [url] or audit [url] for [business name]
Crawl limit	50 pages max
Parallel steps	6 (dispatched after crawl completes)
Output files	src/content/reports/[client-slug]-internal-[date].md + [client-slug]-prospect-[date].md
Packs required	scraping, google-apis, performance, seo, ai-visibility, client-ops
Health check	node .claude/lib/doctor.js

Reference Index

I want to…	File
Run the full 7-step audit pipeline	reference/audit-pipeline.md
Generate internal + client-facing reports	reference/dual-output.md
Handle missing or unavailable data	reference/missing-data.md

Usage: Read the reference file matching your current task from the index above. Each file is self-contained with code examples and inline gotchas.

---

Announcement

┏━ 🔍 prospecting ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┓ ┃ Auditing prospect: [url] for [business name] ┃ ┗━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┛