SEO Inside VS Code: Live Search Data in Copilot

You are in VS Code with Copilot Chat open in Agent mode, shipping a feature. Copilot just scaffolded a marketing page, you are writing the H1 to sit on top of it, and the doubt lands. Is anyone actually searching for the phrase you just typed, or are you optimizing a page for a keyword nobody uses? The old reflex was to break flow, open a browser, log into a $129 dashboard, run one lookup, and lose your place on the way back.

There is a cleaner move now. VS Code’s Model Context Protocol support went generally available in 2025, so Copilot can call external tools mid-chat the same way it already reaches GitHub or your database. Point it at a live SEO data source and you ask that keyword question right in the chat panel, get a real volume and difficulty number back in seconds, and stay in the file you were already editing.

This guide covers how to wire SEO data into VS Code, the two settings that trip everyone up, what you can actually ask once it is connected, and the honest limits of working this way. I build solo, I write my own copy, and I run this workflow every day, so this is how it actually goes, not a spec sheet.

Why do SEO inside VS Code

The pitch is not that VS Code is secretly an SEO tool. It is that the cost of asking an SEO question should be close to zero, and for most builders it still is not.

Most people who do their own SEO use a sliver of what a dashboard offers, and only now and then. You want to know if a keyword is winnable before you write the page. You want a quick read on a rival’s traffic. You want to check that the headline you just wrote matches how people actually search. None of that needs a forty-tab research session. It needs one number, fast, without leaving what you are doing.

The Model Context Protocol, or MCP, is what makes the shortcut possible. It is an open standard, introduced by Anthropic in late 2024, that lets AI agents call outside tools and data through one common interface. In VS Code the client is GitHub Copilot Chat, and once you connect an SEO server it can call those tools mid-conversation, the same way it already reaches your file system or a database. SEO data becomes just another source Copilot can pull from.

The payoff is flow. When the answer to “is this keyword soft enough to rank” lives in the same window as your code and copy, you ask it. When it lives behind a login two tabs away, you guess instead, and guessing is how pages end up optimized for phrases nobody types. If you want the longer argument for why asking the agent beats opening the dashboard, the SEO MCP server explainer lays out the full case.

How do you connect SEO data to VS Code?

There are two config files, and both take a couple of minutes. Workspace config lives at .vscode/mcp.json and is good for servers you want to commit for a team. User config syncs across your machines via Settings Sync, which is where a personal server holding your token belongs. Open the Command Palette and run MCP: Open User Configuration or MCP: Open Workspace Folder Configuration to jump straight to either file.

Here is the whole entry for a hosted server like ContextBolt SEO, which hands you one URL on signup.

{
  "servers": {
    "contextbolt-seo": {
      "type": "http",
      "url": "https://seo.contextbolt.app/mcp/YOUR_TOKEN"
    }
  }
}

Two details are worth tattooing on the back of your hand, because they are the reason a fresh server so often “does not work” in VS Code.

First, the root key is servers, not mcpServers. Every other client, Cursor, Windsurf, Claude Desktop, uses mcpServers, so if you paste a working block out of one of their guides, VS Code silently loads nothing. VS Code also requires an explicit type field on every server, where the others infer it. Set "type": "http" for a remote server like this one.

Second, MCP tools only run in Agent mode. Open Copilot Chat, click the mode dropdown in the chat box, and switch from Ask to Agent. In Ask mode the tools simply never fire, and this is the single most common reason people think the server is broken when the config is fine. For the record, the field names genuinely differ across clients too. VS Code uses servers, Cursor and Codex use url under mcpServers, Windsurf and Antigravity want serverUrl, Gemini CLI wants httpUrl. Same server, different wrapper.

Once the file is saved and you are in Agent mode, ask Copilot “what SEO tools do you have?” and you should see them listed. One caution before you commit anything. Your MCP URL carries a personal token, so keep it in the user config, not a committed .vscode/mcp.json, or reference it through the inputs prompt VS Code supports. If MCP itself is new to you, What Is MCP? is the short primer.

What can you actually ask once it’s connected?

This is where it stops being a config exercise and starts earning its keep. You type a question, Copilot picks the tool, the data comes back in the chat. A few of the prompts I lean on most.

• Keyword reality checks: “What keywords should my new product page target?” Copilot calls keyword research and returns related terms with volume, difficulty, and intent, so you plan the page while you build it.
• Difficulty and SERP in one shot: “How hard is ‘ai code review’ to rank for, and who’s in the top 10?” It chains a difficulty score and the live page-one results, so you see both the bar and the field before you commit.
• Domain teardowns: “What does supabase.com rank for, and who else competes?” Copilot runs the ranked-keywords lookup, then competitor keywords, and hands back a content map plus the rivals chasing the same terms.
• Backlink gap mining: “Who links to my main competitor but not to me?” You get a spam-scored prospect list, a ready-made outreach shortlist. Backlink lookups cost more, so you run them on purpose.
• Winnability triage: “Of these five keywords, which can a new domain realistically rank for?” One prompt, a ranked answer, no spreadsheet.
• On-page audits: “Audit my site and list the worst pages first.” You get a graded fix-list, and because Copilot edits files in Agent mode, you can say fix the top three and re-audit without leaving the workspace.

The thread through all of these is plain language. You are not learning a menu or remembering which report lives where. You ask the way you would ask a colleague, and Copilot translates it into the right lookup, then chains the next one when the question needs two. “Find five low-difficulty keywords for my features page, check who ranks for each, and note the gaps” is one sentence that fires several tools and writes the whole session to disk.

Where ContextBolt SEO fits

ContextBolt SEO is the one I make, and it exists for exactly this workflow. It is a hosted MCP server that wraps Ahrefs-grade data behind a single URL. You paste that URL into your VS Code config as one http server, and Copilot has nine tools to reach for. Six are research tools and cost one lookup each: keyword research, keyword difficulty, SERP overview, domain overview, ranked keywords, and competitor keywords. Three are backlink tools and cost three lookups each: backlink overview, referring domains, and competitor gap. It runs $35 a month for 1,000 lookups, a quarter of the entry price of the big suites, with no account to provision and no deposit to fund. You subscribe, you get a URL, you work. The free Google Search Console tools and the automatic memory never touch your lookup budget.

Two things make it fit VS Code specifically rather than just any MCP client. First, every tool returns a digested answer with a verdict line, not a wall of raw JSON, so the chat panel stays readable while you code. Second, because Copilot in Agent mode can edit files, every lookup also mirrors to a ./seo-findings/ folder in your workspace as markdown, one clean file per keyword or domain, plus an index. Your SEO research lives in the repo next to the feature it informed. You can grep it, commit it, diff it next month, or open it in Obsidian. Ask about the same keyword a week later and the answer leads with what changed since you last looked, no extra prompt needed. Keep the server in your user config so it follows you across machines via Settings Sync, and the full tool guide walks through every lookup.

There is one more move that only works because this runs inside a coding agent. Say “audit my site” and ContextBolt SEO grades the mechanical health of your pages against a strict ruleset, title and meta lengths, a single H1, thin content, broken links, canonical tags, then hands back its own 0-100 score with the fix-list worst first. Because Copilot edits the files it just checked, you say “fix the top three” and it rewrites the pages in the same session, then re-audits and leads with the delta, health 80 to 91. Ahrefs reports the problems. ContextBolt hands them to the thing that can fix them. Find it, then fix it.

The honest framing the product holds to applies here too. This is Ahrefs-grade data, not the same numbers as Ahrefs. It sits on DataForSEO estimates, which are decision-useful and directionally accurate, the same class of data inside many dashboards you already trust. For deciding what to build and what to write, that is more than enough. It is not the tool for a client report that has to match Ahrefs to the decimal, and I would rather say that plainly than let you find out later.

The honest limits

Doing SEO inside VS Code does not replace everything a suite does, and pretending it did would be the kind of overclaim this blog tries to avoid.

You do not get a visual dashboard. If you think in charts and like clicking around a keyword explorer, the chat panel will feel bare at first. You do not get scheduled rank tracking. An MCP server answers when you ask, it does not watch your positions every day and email you when one moves. If daily rank monitoring is central to your job, keep a tool that does it. There is also a VS Code-specific catch worth naming. Copilot tool calls need an active GitHub Copilot subscription, even on the free tier, and they only run in Agent mode, so this is not a workaround for having no Copilot at all. And Copilot can be confidently wrong about strategy even when the underlying numbers are right, so treat its SEO advice the way you treat its code, fast, useful, and in need of a human read before it ships.

There is a usage shape to respect too. A metered server has a monthly lookup budget, so you spend it on real questions, not idle curiosity. In practice that is a feature, because it nudges you toward the lookups that actually move a decision. But it is worth knowing going in.

What you get back is the thing that matters most when you build solo and time is the scarce resource. The answer arrives where you already are, in the seconds you would otherwise burn on a context switch, at a price that does not make you wince. For most builders doing their own SEO, that trade is an easy yes. The keyword you were nervous about might be wide open, and now you can find out without ever leaving VS Code.