Skip to main content
The Stophy skills teach a coding agent which Stophy command fits a task, how to run it and how to read the result. Each skill is a SKILL.md file in the skills folder of the stophydotdev/stophy repository. A skill is instructions only. It allows the agent to run stophy ... and npx @stophy/cli ... shell commands. The skills run on the Stophy CLI. If stophy is missing or too old, they fall back to npx @stophy/cli@latest.

Install

For any agent that supports the Agent Skills format, install every skill.
To install only the skills you need, name them.
stophy init runs this command for you, and also logs you in and adds the MCP server. See Install the CLI.

Install as a plugin

The repository is also a plugin for three coding agents. Each manifest sits at the repository root and names the hosted MCP server. For Claude Code, the marketplace entry points at the repository root (source: ./), so the plugin brings the skills in skills/ and the MCP server named in .mcp.json, https://mcp.stophy.dev/mcp. You sign in with your browser the first time. See Browser sign-in. skills.sh.json groups the skills for skills.sh.

Sign in

The skills call the CLI, which needs a login for everything except search, scrape, map, parse and the parsers that need no API key. Those work up to 10 calls a minute and 50 calls a day per IP address. Log in with your browser, or set STOPHY_API_KEY.

What each skill is for

Start with stophy. It names the skill and the command for your task. The others cover one kind of task or one kind of site.

Start here and tasks

Build with Stophy

Workflows

Site data

Check that it loaded

Start a new agent session and ask this.
A loaded skill answers with stophy scrape "https://bun.com/docs". If the agent answers with a plain HTTP fetch instead, check that each skill sits in a folder your agent reads, as <name>/SKILL.md, and restart the agent session.