Traction scoreGitHub stars can be faked, so popularity alone can be misleading. Traction Score looks for broader signs of recent attention, adoption, and active maintenance.
This collection provides structured frameworks for growth marketers and analysts to execute conversion experiments with statistical rigor. It transforms the process of A/B testing from intuitive guessing into a disciplined practice of hypothesis testing and sample size calculation.
By integrating specific methodologies for split testing and multivariate analysis, the skill set helps users determine test duration and statistical significance to ensure data-driven decision making.
Compatible AgentsThe repository documents support for these agents. The skills may also work with other agents that can load SKILL.md files, but they may need some setup or small changes.
Follow the documented setup, then try a first task.
Install the complete OpenClaudia bundle
Installs: All 78 OpenClaudia marketing skills. skillPaths is a bounded selection emphasizing the starter tasks; the command installs the complete documented bundle. · Claude Code
Before you start
Claude Code installed and authenticated.
For github-stars: the gh CLI installed and authenticated, plus Python 3.9 or newer.
For similarweb-traffic: Node.js with built-in fetch support.
For geo-difficulty: Bash for the bundled script, or curl for the documented direct API call.
Run this terminal command to install the complete bundle.
Installs: The OpenClaudia repository bundle. skillPaths is a bounded selection emphasizing the starter tasks; other repository skills may also be installed. · Claude Code
Before you start
Claude Code installed and authenticated.
The skills CLI must be runnable through npx.
For github-stars: the gh CLI installed and authenticated, plus Python 3.9 or newer.
For similarweb-traffic: Node.js with built-in fetch support.
For geo-difficulty: Bash for the bundled script, or curl for the documented direct API call.
Run this terminal command to add the repository bundle.
A client domain for the client-relative verdict; this input is optional if only absolute difficulty is needed.
Use the geo-difficulty skill to score "<keyword>" for the client domain <domain>. Replace <keyword> with the search query and <domain> with the site to evaluate. Report the absolute difficulty, client verdict, KD, median competitor-page UR, targeting percentage, AI Overview status, and principal competitors. Treat the result as a current snapshot because SERPs and UR can change.
Node.js with built-in fetch support; no API key or npm install is required.
Use the similarweb-traffic skill to produce a concise traffic snapshot for <domain>. Replace <domain> with a bare domain or URL. Include estimated monthly visits, recent visit trend, engagement metrics, traffic-source mix, top countries, top keywords, and available ranks. Clearly label the figures as SimilarWeb estimates and report all-zero results as no SimilarWeb data.
An optional IANA timezone; the documented default is America/Los_Angeles.
The gh CLI installed and authenticated.
Python 3.9 or newer.
Use the github-stars skill to show daily and hourly star growth for <owner/repo> in <IANA timezone>, using the default 14-day daily and 2-day hourly windows. Replace <owner/repo> with the exact GitHub repository and <IANA timezone> with a timezone such as UTC or America/New_York. Mention the 40,000-star timestamp cap if applicable and explain that the chart reflects currently retained stars rather than gross starring events.
Claude
Codex
Cursor
Gemini CLI
Windsurf
Grok