
rankizz
io.github.Rankizz/rankizz
SEO research and tracking: keywords, SERPs, backlinks, site audits, Search Console and GA4.
Documentation
Rankizz Agent Skills
Portable SEO and audience-research workflows for Claude, Codex, ChatGPT, and other agents that support the Agent Skills format.
Each skill is an independent folder with a public SKILL.md. Some workflows can run entirely with local files and network tools. Others can optionally use Rankizz MCP or the Rankizz REST API for live search, backlink, analytics, and audit data.
Rankizz plugin (MCP + skills)
plugins/rankizz is the Rankizz connector plugin: it connects your agent to the hosted Rankizz MCP server (https://www.rankizz.com/mcp, OAuth) and bundles the skills that use it. Install it in Claude Code with:
/plugin marketplace add Rankizz/skills
/plugin install rankizz@rankizz-skills
The root skills/ folder and the rankizz-seo plugin stay as before for agents that only want portable skills.
Skills
| Skill | Purpose | Credit usage |
|---|---|---|
rankizz-seo-skills | Set up Rankizz and install the right specialist skills for the task | Optional |
rankizz-setup-seo-project | Create reusable SEO project context | Optional |
rankizz-audit-site-cloud | Run a hosted Rankizz site audit | Required |
rankizz-audit-site-local | Audit with the agent's local resources | None |
rankizz-audit-on-page-seo | Check the complete on-page SEO implementation | Optional |
rankizz-research-keywords | Find and prioritize keyword opportunities | Optional |
rankizz-cluster-keywords | Turn keyword lists into a page map | Optional |
rankizz-map-competitive-landscape | Map the domains winning a search market | Optional |
rankizz-analyze-competitor | Analyze one competitor's organic footprint | Optional |
rankizz-prospect-links | Find and qualify relevant link prospects | Optional |
rankizz-build-seo-roadmap | Build an evidence-based 1–30 day SEO roadmap | Required |
rankizz-write-human-blog-post | Research and write a useful post in a natural voice | Optional |
rankizz-write-tool-listicle | Create an evidence-backed best-tools list | Optional |
rankizz-write-alternatives-page | Write a fair product alternatives page | Optional |
rankizz-write-comparison-page | Compare two products using the same evidence standard | Optional |
rankizz-write-competitor-review-pricing | Explain a competitor's product, fit, and current pricing | Optional |
rankizz-write-reddit-post | Write a useful, subreddit-native post | Optional |
rankizz-find-reddit-reply-opportunities | Research threads and save transparent reply drafts | Required |
rankizz-write-linkedin-post | Research current posts and draft a knowledge-first LinkedIn post | Required |
rankizz-find-outreach-emails | Find public, role-relevant outreach contacts with source evidence | None |
rankizz-write-backlink-outreach-email | Write a credible page-specific backlink pitch | Optional |
Install
With the skills CLI, which supports Claude Code, Codex, Cursor, and many other agents:
npx skills add Rankizz/skills # choose interactively
npx skills add Rankizz/skills --skill rankizz-research-keywords
From npm, into Claude Code (.claude/skills) or Codex (.agents/skills):
npx @rankizz/skills # all skills, current project
npx @rankizz/skills rankizz-research-keywords --global # one skill, personal
npx @rankizz/skills --agent codex
Full instructions for every client are at rankizz.com/docs/skills. Upload-based clients (Claude.ai, ChatGPT) can still use a ZIP built with python3 scripts/build_packages.py.
Rankizz access
Remote Rankizz workflows use the current Streamable HTTP MCP endpoint:
https://www.rankizz.com/mcp
Interactive clients should use OAuth. Headless clients can create a personal key from Rankizz API access. MCP and API calls require a balance of at least 10 credits. A successful no-cost operation does not deduct credits.
Build packages
Run:
python3 scripts/build_packages.py
The script validates the catalog and creates one self-contained ZIP per skill in dist/.
No installation packages available.