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.
Studies an app’s features and flows, then rebuilds them with an original identity. Its stated boundary is clean-room functionality, not copying the source code, branding, or content.
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.
I’m using
Install through Claude
Installs: The Replica skill pack of eleven skills; the supplied evidence does not establish individual file mappings for this installation workflow. · Claude
Before you start
Python 3.8 or newer must already be installed to use the pack's Python tools; no pip packages are required.
Installs: All replica-* skill folders copied globally into ~/.claude/skills/. The listed paths are a bounded selection from the eleven-skill pack. · Claude Code
Before you start
Python 3.8 or newer must already be installed to use the pack's Python tools; no pip packages are required.
Run these commands in a terminal, in order, to clone the repository and copy its skill folders into the global Claude skills directory. Folder installation exposes plain skill names such as /replica-recon.
Installs: All replica-* skill folders copied into the target repository's .claude/skills/. The listed paths are a bounded selection from the eleven-skill pack. · Claude Code
Before you start
Python 3.8 or newer must already be installed to use the pack's Python tools; no pip packages are required.
Run this command in a terminal to obtain the repository.
Python 3.8 or newer for the supplied contrast checker.
Make sure replica-design is installed in your agent before trying this task.
Use replica-design to draft the initial design tokens from replica/recon.md and the reference screenshots in replica/screens/. Identify colour roles, type and spacing scales, and layout measurements. Use a neutral accent placeholder and open font alternatives. Produce replica/design/tokens.json and tokens.css, include the text/background pairs, and check their AA contrast.
The core flow to prioritize and enough information about the clone to distinguish automated and manual checks.
Make sure replica-test is installed in your agent before trying this task.
Use replica-test to create a focused test plan for one core flow from replica/recon.md. Write its happy path, applicable edge cases, and negative cases with numbered case IDs in replica/test-plan.md. Keep the target limited to my clone and identify which cases can be automated and which need a manual checklist.
A current replica/features.csv with feature priorities and clone statuses.
Python 3.8 or newer for the supplied parity tool.
Make sure replica-diff is installed in your agent before trying this task.
Use replica-diff to check feature parity from replica/features.csv. First check that each clone status is yes, partial with a note, no, or skip with a reason. Calculate the weighted feature score, summarize must-haves completed, and give me the top five missing features in build order. Keep this first pass focused on features.