How We Used AI to Find Gaps in Our Content Strategy

We had 119 blog posts and no idea which topics we were missing. So we used Claude Code, connected to AgentPatch, to audit our content strategy against real search trend data. The process took about 30 minutes. Here’s what we did and what we found.

The Problem

AgentPatch is a tool marketplace for AI agents. Our blog covers topics related to AI agents, MCP, tool use, and developer workflows. We’ve been publishing consistently, but our approach to topic selection was informal: write about what seems relevant, publish, move on.

That approach works until it doesn’t. We had no systematic way to know whether our 119 posts covered the topics people were searching for, or whether we had blind spots where demand was growing and we had zero coverage.

The Process

We opened Claude Code with AgentPatch connected and ran a structured research session. The goal: find high-growth keywords in our space that we had no blog posts targeting.

Step 1: Keyword research. We asked Claude Code to use the Google Trends tool (available on the AgentPatch marketplace) to research 45 keywords across 9 queries. Each query pulled trend data for 5 related keywords over the past 12 months. The keywords covered AI agents, developer tools, automation, and related topics.

Step 2: Growth analysis. For each keyword, we looked at the growth trajectory: where it started 12 months ago and where it is now. We were looking for keywords with strong upward trends, not static or declining search interest.

Step 3: Cross-reference with existing content. We had Claude Code scan our 119 existing blog posts and check whether each high-growth keyword was covered by at least one post. A “covered” keyword had a dedicated post targeting it. An “uncovered” keyword had zero posts.

Step 4: Gap identification. We filtered for keywords that met two criteria: strong growth on Google Trends and zero existing coverage in our blog.

What the Data Showed

The Google Trends data revealed clear patterns. Some keywords we expected to be growing were flat. Others we hadn’t considered were surging.

Here are some of the growth numbers that stood out:

  • “Vibe coding”: grew 23x over 12 months (from 4 to 91 on Google Trends)
  • “AI code review”: grew 50x (from 1 to 50)
  • “AI phone agent”: grew 33x (from 3 to 100)
  • “AI terminal”: grew from 9 to 88
  • “Claude CLI”: went from 0 to 69
  • “OpenAI Agents SDK”: appeared in March 2025, reached 52
  • “AI API”: grew from 15 to 100
  • “n8n AI”: grew 5x
  • “AI content strategy”: steady growth to 72
  • “AI workflow automation”: grew from 22 to 91

We also found keywords with strong volume but no growth (already mature topics) and keywords with growth but low absolute volume (too niche to prioritize).

What We Were Missing

Our existing 119 posts had strong coverage in three areas:

  • Claude Code workflows. Multiple posts on tool chaining, email, image generation, and research with Claude Code.
  • MCP ecosystem. Posts on MCP servers, the marketplace concept, and how MCP works.
  • Email automation. A deep bench of posts on agent email, including setup guides for every supported client.

But we had zero posts targeting 12 high-growth keywords:

  1. Vibe coding
  2. AI code review
  3. AI phone agent
  4. AI terminal / Claude CLI
  5. OpenAI Agents SDK
  6. AI API (general)
  7. n8n AI / AI workflow automation
  8. AI content strategy / AI SEO
  9. AI customer support
  10. AI lead generation
  11. AI research assistant
  12. Free AI image generator

That’s 12 topics where search demand was growing fast and we had nothing published. Each one represented people searching for something related to our product, finding nothing from us.

What We Did About It

We wrote 13 new blog posts targeting these gaps. This post is one of them.

Each post was planned based on the trend data: the keyword, the growth trajectory, the search intent behind it, and how AgentPatch relates to the topic. Some posts are guides (how to use Claude Code as a terminal). Some are comparisons (n8n vs AI agents). Some are listicles (best AI APIs). The format matches what the searcher is looking for.

The entire analysis, from raw keyword research to a prioritized list of posts with outlines, happened in a single Claude Code session with the Google Trends tool. The agent did the research, we made the editorial decisions.

You Can Do This Too

The Google Trends tool is available on the AgentPatch marketplace. Any AI agent connected to AgentPatch can query trend data, analyze growth curves, and compare keywords. You don’t need an SEO platform subscription or a marketing team.

The process works for any content operation:

  1. List the keywords relevant to your business.
  2. Pull trend data for each one.
  3. Cross-reference against your existing content.
  4. Find the gaps where demand is growing and you have no coverage.
  5. Write posts targeting those gaps.

The hard part used to be steps 1 through 4. With an AI agent and the right tools, those steps take minutes instead of days.

Setup

Connect AgentPatch to your AI agent to get access to the tools:

Install the AgentPatch CLI (zero dependencies, Python 3.10+):

pip install agentpatch

Set your API key:

export AGENTPATCH_API_KEY=your_api_key

Then use it:

ap search "web search"
ap run agentpatch google-search --input '{"query": "test"}'

Get your API key from the AgentPatch dashboard.

Claude Code

Run this command to add AgentPatch as an MCP server:

claude mcp add -s user --transport http agentpatch https://agentpatch.ai/mcp \
  --header "Authorization: Bearer YOUR_API_KEY"

Replace YOUR_API_KEY with your actual key from the AgentPatch dashboard. Claude Code discovers all AgentPatch tools automatically.

OpenClaw

Add AgentPatch to ~/.openclaw/openclaw.json:

{
  "mcp": {
    "servers": {
      "agentpatch": {
        "transport": "streamable-http",
        "url": "https://agentpatch.ai/mcp",
        "headers": {
          "Authorization": "Bearer YOUR_API_KEY"
        }
      }
    }
  }
}

Replace YOUR_API_KEY with your actual key from the AgentPatch dashboard. Restart OpenClaw and it discovers all AgentPatch tools automatically.

Agents Do More Than Code

This project changed how we think about AI agents. We use Claude Code every day for development work: writing code, debugging, reviewing PRs. Using it for content strategy felt different. The agent wasn’t writing code. It was doing market research, analyzing data, and helping us make business decisions.

The insight is that agents with the right tools can do useful strategic work. The Google Trends tool costs a few credits per query. A full keyword research session costs less than a dollar. The output was a prioritized content plan based on real demand data. That’s a lot of value from a 30-minute session.

AI agents aren’t limited to coding tasks. Give them tools that access real-world data, and they become useful for strategy, research, analysis, and planning. The tools define the boundaries of what an agent can do. Wider tool access means a wider range of useful work.

Wrapping Up

We used our own product to find gaps in our content strategy. The process worked well enough that we’re planning to run it monthly. If you’re producing content and want to know what you’re missing, connect Claude Code to agentpatch.ai, run the same analysis, and see what the data tells you.

← Back to blog