Give your agent
a full moveset
We build the result first, then turn it into a workflow your agent can run. Works in plain language.
Featured skills
View all →Designed and written in-house; each product page lists exactly which agents it has been tested on. Thumbnails are samples, labelled as demo data.
Find by task
Not sure what it's called? Start from the job.
Free skills
View all →Install, run, keep what it makes. No card.
ship with new releases
Every skill starts
as a finished result
We do the job by hand first. Only when the result holds up do we write the workflow your agent follows.
- Original methods, templates and scripts
- Every version pinned by SHA-256
- Maintained by the team that built it
Do the real job once, define what good output looks like.
Method, templates and scripts packed as one SKILL.md.
Only tested agents and real outputs on the page.
Platforms change, models change; we update.
No terminal?
Let your agent do it
Paste the prompt to your agent. It reads the guide, asks you to authorise in the browser, then installs and verifies.
Read https://moveset.io/install/agent.md and follow it to install moveset, then recommend the right skills for me. I run a Shopify store and want to study competitors.Prefer a terminal? $ npx moveset add <skill>
What we won't claim
Every finding carries an evidence label; unknowns say unknown.
Network, commands, files — known before you buy.
Won't install or run in the stated setup within 14 days? Full refund.
We show the work; we don't promise revenue.
Buy one, or have it done
Alipay, WeChat Pay or card. One-time, includes 12 months of updates.
Pro
- Multi-store compare, scorecards
- Browser-authorised install + ZIP
- 14-day refund if it fails
Human teardown
- 1 benchmark vs your store
- Report + 30-min call
- Pro fully credited
Before you start
What you need to know for your first skill.
What exactly is a skill?
No Claude — does it work?
Any API costs?
How do updates work?
Need a skill
we don't have?
Describe the job you repeat every week. We read every request and reply whether we can build it, and roughly when.