LeaderboardEvery site, rankedFree toolsChecks, no sign-upAgentsYour agent fixes itDocsThe API and MCP

Agent skills

Folders of instructions your agent loads from your site’s own repository. Each does one job, and checks its work on the live site with mttr.

  1. 1. Add the skillsnpx skills add subinternal/mttr
  2. 2. Tell your agentCheck yoursite.com with mttr and fix what’s worth the most.

The skills

Start with mttr. The others each do one job.

  1. mttr

    v1.2.1

    Teaches your agent to use mttr end to end: connect, check a site or a page, read the brief, fix in order of worth, propose or ship as you allow, then read again and compare.

    For
    Anyone with a site and an agent, whatever the site is built with.
    Ask
    Check example.com with mttr and fix the top three things, then show me what changed.

    62MCP tools in its reference

    See the skill

  2. Shop readyNew

    v1.2.0

    Fixes the Product data in a shop's own product templates so it states what the page shows, and checks each live page with mttr before and after.

    For
    A shop’s developer, in the theme or storefront repository.
    Ask
    Fix the product schema in our Shopify theme so every product page passes Google’s merchant listing fields.

    16fields checked on each page

    See the skillRead the Universal Cart guide

  3. Page system

    v1.3.1

    Builds a site’s page templates and shared components so crawlers and assistants read every page well, and proves each pattern with mttr. About structure and spacing, not color, type or brand.

    For
    Someone building a site’s templates, or rebuilding them.
    Ask
    Rework our page templates so every page has one h1, the right structured data and a breadcrumb, then check them with mttr.

    38mttr rules its patterns answer

    See the skill

  4. llms.txt

    v1.1.2

    Adds a /llms.txt to a site, a plain text outline for AI assistants built from the site’s own pages and facts, with tests that catch a missing page, and checks it with mttr.

    For
    A site owner who wants assistants pointed at the right pages.
    Ask
    Add an llms.txt to this site from our sitemap, with a test that fails when a page is missing.

    6tests it adds to the site

    See the skill

  5. Livecards

    v1.2.0

    Makes one page of a site a livecard, a page that runs inside a post on X: the player card tags, framing for X on that route only, what X’s crawler is really served, and an embed mode for scrolling, links and overlays, checked live with mttr.

    For
    A site’s developer who wants one page to work inside a post.
    Ask
    Make our /demo page run inside a post on X with a player card, and check what X’s crawler gets.

    9card tags it writes

    See the skill

  6. Rank a page

    v1.2.1

    Runs a campaign to move one page up for one search toward an aim, through five gates in order, each opened only by a proof from a read of the live site, with a weekly read-out and the campaign’s time to rank.

    For
    A site’s owner or developer with one page that should rank higher.
    Ask
    Rank our /hiking-boots page for “waterproof hiking boots”, aiming for top 3, and keep a weekly read-out.

    5gates, each opened by a proof

    See the skill

Install in any agent

The line above adds every skill. Where each agent looks for them:

  • Claude Code

    Loads skills from .claude/skills/ in the repository, for everyone who works in it, or ~/.claude/skills/ for every project on your machine.

  • Codex

    Loads skills from .agents/skills/ in the repository and its parent folders, or ~/.agents/skills/ for every project.

  • Cursor

    Loads skills from .cursor/skills/ or .agents/skills/ in the repository, or ~/.cursor/skills/ for every project. It also reads .claude/skills/, so one copy there serves both.

  • Any other agent

    An agent that does not load skills can still follow the file. Save the folder in the repository and point its standing instructions at it, with a line in AGENTS.md, which many agents read.

It needs Node. To add one skill alone, or on Windows without Node, each skill’s page has its own commands. Every skill is listed as data at /skills.json.

How a skill works

  1. 1Installed in the repository

    One command puts the folder where your agent looks, so everyone who works on the site has it.

    .claude/skills/mttr/SKILL.md
  2. 2Loaded when the work matches

    Your agent reads the skill when a request matches its description, or when you call it by name.

    /mttr-shop-ready
  3. 3Checked by mttr

    Each skill checks the live pages with mttr before and after a change, and says what is left for you.

    mttr_rescan, then mttr_changes

Every skill is in the open Agent Skills format, served as it is in mttr’s repository, with its version under metadata in SKILL.md and its changes in CHANGELOG.md. A new skill is listed here the day it ships.