More screenshots
Works with the AI Agents you already use
Growth Stack: SEO and Marketing Skills in One Pack
**How it works.** You ask in your own words ("audit this page", "rewrite my title and description", "review my pricing page").
$14.99
Growth Stack: SEO and Marketing Skills in One Pack
Example session with this skill installed
Audit my homepage, fix the biggest gaps in my own voice, and show me the score before and after.
- Read your context and instructions
- Compiled the growth stack
The page is a fictional bike shop fixture served locally. The pack's worked example has every command and every output.
Step 1: the audit (page_audit.py, real output)
GROWTH SCORE: 23/100
Crawl and index 13/25
On-page SEO 10/30
Structured data 0/10
AI-search readiness 0/15
Conversion 0/20
1. [HIGH] cta -12: verb-led CTA (-6); specific label (-3); largest form has 5 fields or fewer (rule of thumb) (-3)
2. [HIGH] snippet -11: title 30-60 characters (-2); has a meta description (-3); description 70-160 characters (-2); og:title, description, image and twitter:card (-4)
3. [HIGH] schema -10: has JSON-LD (-4); all blocks parse (-3); common type has required properties (-3)
Step 2: loop 1 fixes the top three gaps (button and form, title, description and share tags, business markup):
score 23 -> 56
+ snippet title 30-60 characters 0 -> 2
+ snippet has a meta description 0 -> 3
+ snippet description 70-160 characters 0 -> 2
+ snippet og:title, description, image and twitter:card 0 -> 4
+ schema has JSON-LD 0 -> 4
+ schema all blocks parse 0 -> 3
+ schema common type has required properties 0 -> 3
+ cta verb-led CTA 0 -> 6
+ cta specific label 0 -> 3
+ cta largest form has 5 fields or fewer (rule of thumb) 0 -> 3
Step 3: loop 2 fixes the next three (mobile and language basics, question headings with short answers, contact and policy links):
score 56 -> 72
+ basics viewport meta 0 -> 2
+ basics html lang 0 -> 1
+ answers 2+ question headings 0 -> 4
+ answers short direct answers (heuristic) 0 -> 4
+ trust contact details 0 -> 3
+ trust policy links 0 -> 2
Step 4: the site probe (site_probe.py) on the same site found a sitemap URL that redirects and no Sitemap line in robots.txt:
[MEDIUM] sitemap: http://127.0.0.1:8765/services: the sitemap lists a URL that redirects
Fix: List the final URL.
[LOW] robots.txt: no Sitemap line
Fix: Add 'Sitemap: <absolute URL>' to robots.txt.
After loop 2 the score is 72/100, and the probe reports no findings. Still lost: 28 points. Six are hosting (https and server headers, because the fixture runs on plain Python), three are review proof, because I will not write a testimonial the owner does not have, and the worked example gives the reason for the other 19.
Connects securely to your tools. The creator never sees your data.
About this skill
Most SEO output is a long list with no order and no proof. Most marketing advice never looks at the page. Growth Stack does both in one loop: it measures one page, ranks the gaps, routes each to the skill that fixes it, and measures again so you can see exactly which points moved.
How it works. You ask in your own words ("audit this page", "rewrite my title and description", "review my pricing page"). The agent picks the matching skill from its description. When a skill needs numbers it writes a small standard-library Python script to a temporary file, runs it, and deletes it. page_audit.py gives a Growth Score out of 100 across crawl and index, on-page SEO, structured data, AI-search readiness and conversion. Every one of the 100 points is named, and the lost ones come with the evidence, why it matters and the fix. site_probe.py checks robots.txt rules for search crawlers and any other crawler tokens you name, sitemaps and a sample of URLs. text_stats.py measures readability, keywords, passive voice and the you-versus-we balance of your copy, offline.
The fifteen skills. Growth loop, technical SEO audit, on-page SEO, structured data, AI-search readiness, local SEO, keyword clusters, content brief, internal linking, positioning and messaging, landing page copy, email sequence, pricing page review, launch plan and measurement plan. Each has inputs, steps, an output, exit criteria the agent can check, edge cases and rules, and each names its neighbours so the right one loads.
What makes it different. The score is transparent and the fixes are tied to it. A fix that does not recover named points is not a fix. The scripts refuse private addresses, cap size, time and redirects, and check the address before and after connecting.
Where it stops. It reads the HTML a crawler gets first, so JavaScript-only content is not seen. It promises no rankings, traffic, citations or revenue. It will not invent a testimonial, a statistic or a price; missing proof is a marked gap. It tests the main search crawlers by default; you name any AI crawler tokens to test with --agent.
- It reads raw HTML. Pages that build their content in JavaScript will score on what a crawler gets first.
- Heuristic checks (answer length, proof, thin content, CTA wording) are rules of thumb and are labelled as such.
- It does not run Core Web Vitals, backlinks, rankings or analytics. Bring exports and it works from them.
- The skills need your facts. Without a real quote, price or address, they leave a marked gap.
- No AI crawler names are built in. Name the ones you want tested with
--agent.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed today
Requires
Permissions
File Scopes
Network access to the pages you ask it to audit: page_audit.py reads one page, site_probe.py reads that site's robots.txt, sitemaps and a small sample of its pages. No credentials, no accounts. text_stats.py makes no network calls.