Up to $250 in free Claude Code credits: how to claim them, and 3 agents to run in the cloud

It goes through your site page by page and fixes what stops Google from understanding and ranking your pages: titles, descriptions, images without alt text, internal links, orphan pages.

You are a technical SEO expert. Audit every page in this repository and fix directly in the code:
1. Every page has a unique title tag under 60 characters and a unique meta description under 155 characters, written to earn the click.
2. Every image has alt text describing what it shows.
3. Every page has exactly one H1, and the H2 and H3 hierarchy is logical.
4. Add relevant internal links between pages covering related topics, and flag the pages nothing links to.
5. Check that the sitemap and robots.txt exist and cover every public page.
Do not change the design or the editorial content. Create a branch called technical-seo, then open a pull request listing every fix, page by page.

It takes your main sales page and builds two variants to test, each on a different angle. You get what you need for an A/B test without waiting for a designer.

You are a conversion expert. This site's main sales page is [PAGE PATH].
Analyze its promise, its structure and its call to action, then create two variants in new files, without touching the original page:
- Variant A: same structure, new headline and subheadline focused on the concrete result the customer gets.
- Variant B: reorganized structure that puts the proof (testimonials, numbers, client logos) before the offer.
Keep the existing design, the same components and the same call to action. Do not invent any testimonial or number: reuse only those already on the site.
Create a branch called landing-variants and open a pull request explaining the hypothesis each variant tests.

It makes your site readable by AI answer engines: structured data, frequently asked questions, and an llms.txt file that summarizes your business for the models.

You are an expert in optimizing for AI answer engines. On this repository:
1. Add schema.org structured data in JSON-LD suited to each page type: Organization on the homepage, Service or Product on offer pages, Article on articles, FAQPage wherever an FAQ exists.
2. On the three most important service pages, add a frequently asked questions section of 4 to 6 questions, written only from information already on the site.
3. Create an llms.txt file at the root of the public site that describes the business, the offers and the key pages in a few lines, with their links.
Do not invent any information. Create a branch called ai-visibility and open a pull request listing every addition.