Skip to main content
Piloterr
Claude Code logo

MCP & IDE agents

Give Claude Code the Piloterr library in the terminal

Register mcp.piloterr.com with claude mcp add --transport http. Claude Code can scrape, render, and enrich from the same shell you already use for git and tests.

  • One CLI command: claude mcp add --transport http plus --header x-api-key
  • No .mcp.json required, the server is stored in Claude Code’s MCP list
  • Full Piloterr OpenAPI library as tools in a terminal session
  • Fits repo work: CLAUDE.md, local git, and live web data in one loop

At a glance

CLI

claude mcp add

HTTP

transport

190+

library tools

stdio

optional subset

Why connect Claude Code

  • Register once, reuse every session

    claude mcp add writes the HTTP server into Claude Code’s config. Next session, the Piloterr tools are already there.

  • Header flag, not a JSON file

    Pass --header "x-api-key: YOUR_KEY". You do not need Cursor’s .mcp.json unless you also use Cursor on the same repo.

  • Terminal context, not a chat toy

    Claude Code already sees the working tree. Hosted MCP adds blocked-site fetch so it can verify a scraper against the live target.

  • Protocol, not a web page

    https://mcp.piloterr.com/ only completes the handshake when the client sends MCP Accept headers. The CLI does that. A browser does not.

Claude Code jobs that need Piloterr

Shell-first, not an IDE settings walkthrough.

  • Debug a failing scraper in-repo

    Claude Code reads your client, calls the matching library tool, and diffs the live JSON against your parser.

  • CLAUDE.md guardrails

    Tell Claude to prefer Piloterr MCP tools over curl when the domain is on the WAF list.

  • Headless research in tmux

    Leave a session running: news, company enrich, then a commit message. No desktop browser.

  • CI-adjacent checks

    Use the health URL in scripts. Keep tool calls in interactive Claude Code so you do not burn credits in a loop.

Claude Code MCP vs Claude on the web

ApproachDIYPiloterr
claude.ai chatNo repo, no MCP HTTP serverLocal tree plus hosted tools
curl in the terminalYou write every requestClaude picks the tool
Local FastMCP stdioYou maintain a processHosted, full library
Playwright CLICAPTCHA and driftRender tools server-side

Add Piloterr to Claude Code

  1. Step 1

    Create an API key

    Dashboard → API keys. You will pass it on the --header flag.

    Create an API key
  2. Step 2

    Run claude mcp add

    claude mcp add --transport http piloterr https://mcp.piloterr.com/ --header "x-api-key: YOUR_KEY". If the flag name changes, check Claude Code MCP docs.

  3. Step 3

    Start a session

    Open Claude Code in the repo. Confirm the piloterr server is listed and tools appear.

  4. Step 4

    Call a library tool

    Ask it to search Google News or crawl a URL. If nothing shows, hit https://mcp.piloterr.com/health and re-check the header.

Workflow recipes

  • Fix a parser against live Amazon JSON

    Claude Code opens your mapper, calls the Amazon tool, and patches field names that drifted.

  • Write a changelog from Google News

    It pulls recent articles for the product name, then drafts CHANGELOG.md in the repo.

  • Compare two job listings

    Indeed or similar library tools, then a markdown table committed next to the research notes.

  • Smoke-test WebUnlocker from the branch

    You stay in git; Claude Code hits a protected URL and reports status before you merge.

Claude Code vs other Piloterr clients

  • Scenario

    You work in a terminal with Claude Code

    Recommendation: This page

  • Scenario

    You work in Cursor’s GUI

    Recommendation: Cursor MCP

  • Scenario

    You want OpenAI’s coding agent

    Recommendation: Codex MCP

  • Scenario

    Unattended multi-day agent

    Recommendation: Hermes Agent

Claude Code install command

HTTP transport only. Do not use stdio unless you run a custom subset locally.

Bash
claude mcp add --transport http piloterr https://mcp.piloterr.com/ \
  --header "x-api-key: your-x-api-key"

Transparent credit pricing

Pay only for successful requests. Start with +500 credits, then scale with plans from $49/mo.

Premium

$49/mo

18,000 credits

Premium+

$99/mo

40,000 credits

Startup

$249/mo

110,000 credits

Ready to get started?

Your web scraping API is one click away. Start with +500 credits, no infrastructure to set up, no proxies to manage, and no credit card required.

  • +500 credits
  • No credit card required
  • All endpoints included