Analysis updated 2026-05-18
Automatically detect SEO problems like missing descriptions, duplicate titles, and broken canonical links across a large content site.
Draft AI-generated meta description improvements that a human reviews and approves before they go live.
Audit whether AI answer engines like ChatGPT and Perplexity can crawl, read, and cite the site.
Track Google Search Console metrics to catch pages with unusually low click-through rates.
| awizemann/seo-agent | 28998306/magicalcanvas | javlonbek1233/-pixelbattle | |
|---|---|---|---|
| Stars | 36 | 36 | 36 |
| Language | TypeScript | TypeScript | TypeScript |
| Setup difficulty | hard | moderate | — |
| Complexity | 4/5 | 3/5 | — |
| Audience | developer | general | developer |
Figures from each repo's GitHub metadata at analysis time.
Requires a Cloudflare Workers Paid plan for the drafting queue, plus D1, KV, and Queues resources.
Seo-agent is a Cloudflare Worker that watches over a website's search engine optimization and keeps it in good shape as real traffic changes. Instead of running as a one time audit, it works on a daily schedule: it crawls the site's own sitemap, records exactly what search engines and crawlers see on each page, and checks that against a set of fixed rules to spot problems like missing or oddly sized page descriptions, wrong canonical links, broken sitemap entries, duplicate titles, and pages that appear or disappear without a deploy. Beyond classic SEO, the tool also checks how well a site is set up for AI answer engines such as ChatGPT, Claude, Perplexity, and Google AI Overviews, looking at things like whether the site has a working llms.txt file, whether robots.txt blocks AI crawlers, and whether pages actually serve readable content to non JavaScript AI fetchers. When it finds a page with a weak description, it can draft a replacement using Cloudflare's Workers AI, running that job through a queue so a slow AI response never blocks the rest of the scan. Drafts are checked for length and completeness before anyone sees them, and nothing goes live automatically. A person has to approve each proposed change, at which point it is written to a key value store and merged into the site's live pages by a separate small Worker installed on the site itself. Every change is recorded and can be undone. An optional connection to Google Search Console pulls in real search data such as impressions, clicks, and average position, which the tool uses to flag pages whose click through rate looks unusually low and to measure whether an applied change actually helped. Setting it up requires a paid Cloudflare Workers plan, since the drafting queue depends on it, along with Node and the wrangler command line tool. The project is built for TypeScript developers managing large content heavy sites, and it includes a ready made setup script for use with coding assistants. The full README is longer than what was shown.
A Cloudflare Worker that daily crawls a site's sitemap, diagnoses SEO and AI-answer-engine issues, and applies approved fixes automatically.
Mainly TypeScript. The stack also includes TypeScript, Cloudflare Workers, D1.
Setup difficulty is rated hard, with roughly 1h+ to a first successful run.
Mainly developer.
This repo across BitVibe Labs
Verify against the repo before relying on details.