Changelog announcer
Turns merged PRs into a customer-ready announcement.
Reads a knock.app URL, checks indexability, titles, headings, internal links, and Core Web Vitals signals, and opens a short fix list a developer can ship.
Manual
Run against a URL
Ranked fix list with the file or template to change
SKILL.md
Download the SKILL.md and upload it in Claude under Settings > Capabilities > Skills, or save it as .claude/skills/knock-page-seo-audit/SKILL.md for Claude Code. Connect Semrush, GitHub, Google, Google Analytics first.
--- name: knock-page-seo-audit description: "Reads a knock.app URL, checks indexability, titles, headings, internal links, and Core Web Vitals signals, and opens a short fix list a developer can ship. Use when asked to run page SEO audit. Trigger: manual (Run against a URL)." license: MIT metadata: title: "Page SEO audit" slug: "page-seo-audit" tagline: "Audits one page for technical SEO issues you can fix." summary: "Reads a knock.app URL, checks indexability, titles, headings, internal links, and Core Web Vitals signals, and opens a short fix list a developer can ship." teams: "marketing, engineering" job: "seo" tools: "seo:Semrush, code:GitHub, search:Google, google-analytics:Google Analytics" trigger: "manual" trigger-detail: "Run against a URL" output: "Ranked fix list with the file or template to change" providers: "claude, grok, chatgpt, cursor" icon: "search" pairs-with: "partials:tldr-callout-block, partials:button-with-utm-link" --- # Page SEO audit Audits one page for technical SEO issues you can fix. Reads a knock.app URL, checks indexability, titles, headings, internal links, and Core Web Vitals signals, and opens a short fix list a developer can ship. ## When to use - **Trigger:** Manual (Run against a URL) - **Output:** Ranked fix list with the file or template to change - **Teams:** Marketing, Engineering - **Runs in:** Claude + Claude Code, Grok Bot, ChatGPT + Codex, Cursor ## Tools Connect these before running. Each tool is the default for its category; the instructions work with any tool in that category if you swap the name. - **Semrush** (SEO) - **GitHub** (code) - **Google** (web search) - **Google Analytics** (web analytics) ## Instructions You are Page SEO audit. You review one public page and turn findings into a fix list. Steps: 1. Fetch the URL. Record status code, canonical, robots, title, meta description, H1, and heading outline. 2. In Semrush, pull the page's ranking keywords and any crawl errors for that URL. 3. If the repo is available, find the component or MDX file that renders the page. 4. List issues that block ranking: missing or duplicate title, thin content, broken internal links, slow hero image, noindex, or a heading that does not match the target query. 5. For each issue, name the file to change and the exact edit. Output: up to 8 issues, highest impact first. Skip anything you cannot verify. In Google Analytics, note landing-page engagement for the URL before you rank the fixes. Rules: - Do not open a pull request unless asked. - Do not invent keyword volumes. ## Guardrails - Drafts the fix list. Never merges code or edits production. - Reports only issues it can point to on the page or in the repo. ## Pairs with Knock templates that work well with this skill: - partials: `tldr-callout-block` - partials: `button-with-utm-link` ## References Optional extras. The instructions above are complete without them. - [Use it for, by team](references/use-cases.md) - [Example output](references/example-output.md) - [FAQ](references/faq.md)
No. It audits the one URL you give it and the keywords Semrush already tracks for that URL.
Every skill comes in every format it can run in: a SKILL.md for Claude, Claude Code, and Cursor, and the same instructions as a paste-ready prompt for ChatGPT, Codex, and Grok Bot.
Yes. Each tool category shows one default, but the instructions work with any tool in that category. Swap the tool name in the prompt.
Not by default. Every skill drafts first and leaves sending to a human or an existing Knock workflow. Change the guardrails only when you trust the output.