finds.dev← search

// the find

zubair-trabzada/ai-marketing-claude

★ 2,038 · Python · MIT · updated Mar 2026

AI Marketing Suite for Claude Code. 15 marketing skills with parallel subagents — audit any website, generate copy, email sequences, ad campaigns, content calendars, competitive intelligence, and client-ready PDF reports.

A collection of Claude Code SKILL.md files that add marketing-analysis slash commands to Claude Code. You install it and get commands like `/market audit <url>` that orchestrate parallel subagents to score a website across six dimensions. It's aimed at freelancers and agency builders who want to sell AI-powered marketing audits to clients.

The parallel subagent architecture is a legitimate use of Claude Code's multi-agent support — splitting the audit into content, conversion, SEO, competitive, and strategy agents means each one gets a focused context window instead of one bloated prompt. The scoring rubric (weights per category, 0–100 output) gives clients something concrete to react to. The PDF report path via reportlab makes it genuinely deliverable as a client artifact. The uninstaller is a small but real sign of craft — most repos just say 'delete the folder yourself'.

The scores are entirely vibes-based — there's no ground truth, no validation against actual conversion data, and two runs on the same site will give different numbers. The 'parallel subagents' framing is mostly marketing: the agents are just separate prompt files, and whether they actually run in parallel depends entirely on how Claude Code schedules tool calls, which isn't guaranteed or configurable here. The Python scripts (analyze_page.py, competitor_scanner.py) are thin wrappers that do basic HTML fetching — no JS rendering, so single-page apps get analyzed on their skeleton HTML, which makes the SEO and CRO scores meaningless for most modern sites. The install script pipes curl straight to bash and writes into ~/.claude with no integrity check.

View on GitHub → Homepage ↗

// want more like this?

We dig through GitHub every week and send a few repos picked for what you actually care about — each with an honest take like this one.

Get finds in your inbox → Search again →