How to Chain Multiple Tools Together in Claude Code

Claude Code is useful on its own, but it becomes notably more capable when it can pull in live data from multiple sources during a session. AgentPatch gives it access to the full tool marketplace — web search, news, trends, Maps, email — through a single MCP connection. When a task calls for multiple tools, Claude Code uses them together.

Why This Matters

Some coding tasks have a research component before the actual code work. You need to understand the current state of a library, what patterns are common, or what external data looks like. Claude Code handles the research and then keeps going — all in one session.

Multi-tool workflows also come up in data-oriented tasks: pulling real data from multiple sources, cross-referencing results, generating reports. Claude Code can do all of that if the tools are available.

Setup

The AgentPatch CLI is designed for AI agents to use via shell access. Install it, and your agent can discover and invoke any tool on the marketplace.

Install (zero dependencies, Python 3.10+):

pip install agentpatch

Set your API key:

export AGENTPATCH_API_KEY=your_api_key

Example commands your agent will use:

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

Get your API key from the AgentPatch dashboard.

Install the AgentPatch skill — it teaches Claude Code when to use AgentPatch and how to use the CLI:

/plugin marketplace add fullthom/agentpatch-claude-skill
/plugin install agentpatch@agentpatch

MCP Server (Alternative)

If you prefer raw MCP tool access instead of the skill:

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.

Example

You’re building a market research feature for a SaaS app. You want to prototype the output before wiring up the real data pipeline. You tell Claude Code:

“I’m building a market snapshot feature. Run a quick research pass on ‘headless CMS’ — search results, trends, and news — and generate a sample API response I can use to mock the frontend.”

Claude Code runs the research:

Searching Google for "headless CMS"...
Checking Google Trends for "headless CMS"...
Searching Google News for "headless CMS"...

Research complete. Here's a summary and a mock API response:

Market: Contentful, Sanity, and Strapi dominate search results.
Trend: +22% interest over 12 months. "Headless CMS comparison" growing fastest.
News: Contentful raised Series F, Sanity announced v4 in late 2025.

Sample API response for the mock:

{
  "keyword": "headless CMS",
  "trendChange": "+22%",
  "trendPeriod": "12 months",
  "topPlayers": ["Contentful", "Sanity", "Strapi", "Hygraph"],
  "recentNews": [
    { "title": "Contentful raises $175M Series F", "date": "2025-11-12" },
    { "title": "Sanity v4 released with new Studio UI", "date": "2025-10-28" }
  ],
  "sentiment": "positive"
}

Writing this to mocks/market-snapshot.json.

Claude Code researched the market across three tools and produced a mock that’s ready to use. That’s the loop: gather data, understand it, produce something useful.

Wrapping Up

AgentPatch gives Claude Code access to the full tool marketplace in a single connection. Combining tools during a session isn’t special configuration — Claude Code just does it when the task calls for it. See what’s available at agentpatch.ai.

← Back to blog